AmaBhungane CIJ NPC v Minister of Justice and Correctional Services
Privacy rights confirmed by constitutional court and pointed out that state agencies are empowered to “gather, correlate, evaluate and analyse” intelligence and perform other functions such as collecting intelligence showing threats to the country and supply intelligence to other security agencies or governmental departments and liaise with other security agencies and use the intelligence for analysis and other purposes.
Madlanga J (Khampepe J, Majiedt J, Mathopo AJ, Mhlantla J, Theron J, Tshiqi J and Victor AJ concurring)
Jafta J and Mogoeng CJ dissenting in part
Heard on: 25 February 2020
Decided on: 4 February 2021
“ In sum, the confidentiality of lawyer-client communications and journalists’ sources is particularly significant in our constitutional dispensation. There is thus a need that special consideration be given to this fact when interception directions are sought and granted. Plainly there are means that may help minimise this particularly egregious form of intrusion into privacy; particularly egregious because of its impact on other constitutional rights. Some of the foreign examples tell us as much. While reference to them should not be seen as dictation to Parliament, they serve as examples of less restrictive means, which do not subvert the purpose of RICA. RICA is thus unconstitutional to the extent that, when the intended subject of surveillance is a practising lawyer or a journalist, it fails to provide for additional safeguards calculated to minimise the risk of infringement of the confidentiality of practising lawyer and client communications and journalists’ sources.”
Quotations from judgment
Note: Footnotes omitted and emphasis added
 The Constitution proclaims that “[n]ational security must reflect the resolve of South Africans, as individuals and as a nation, to live as equals, to live in peace and harmony, to be free from fear and want and to seek a better life”. It does so against a historical backdrop in which the pursuit of a skewed notion of national security was weaponised and calculated to subvert the dignity of the majority of South Africans. As part of this pursuit, law enforcement involved searches of people, their homes, and their belongings. Over the years, law enforcement evolved to include the surveillance of people, their homes, their movements, and their communications. Today technology enables law enforcement agencies to not only physically – as opposed to electronically – invade the “intimate personal sphere” of people’s lives, but also to maintain and cement its presence there, continuously gathering, retaining and – where deemed necessary – using information.
 At the heart of this matter is the right to privacy, an important constitutional right which, according to this Court, “embraces the right to be free from intrusions and interference by the state and others in one’s personal life”. State intrusions into individuals’ privacy may occur in many and varied ways. In this case, we are called upon to consider intrusions in the context of the surveillance of individuals, including the interception of their private communications, under the Regulation of Interception of Communications and Provision of CommunicationRelated Information Act (RICA). Section 14(d) of the Constitution entrenches the right of everyone “not to have the privacy of their communications infringed”, which is a component of the right to privacy. Whilst RICA prohibits the interception of any communication, it leaves the door wide open for the interception of communications in a variety of ways. It does this by providing that interceptions may be effected as long as that is in accordance with the provisions of RICA. RICA then regulates the circumstances under which communications may be intercepted.
 The first question before us is whether RICA unreasonably and unjustifiably fails to protect the right to privacy and is, therefore, unconstitutional to the extent of this failure. The High Court of South Africa, Gauteng Division, Pretoria, answered this question in the affirmative. It declared that in several respects RICA is deficient in meeting the threshold required by section 36(1) of the Constitution to justify its infringement of the right to privacy, particularly considering the interplay between the privacy right and other constitutional rights protected in sections 16(1), 34 and 35(5). More on these and other constitutional rights later. The declaration of invalidity was suspended for two years to allow Parliament to cure the defects. The High Court granted interim relief by reading-in certain provisions.
 The second question before us is whether there is a legal basis for the state to conduct bulk surveillance. The High Court held that the state’s practice of bulk interception of communications is not authorised by law. This question arises as a result of an appeal by the Minister of State Security against this holding. A subsidiary issue is whether the appeal is properly before this Court.
 The applicants, AmaBhungane Centre for Investigative Journalism NPC, and its managing partner, Mr Stephen Patrick Sole, a journalist, seek confirmation of the High Court’s declaration of invalidity. They support the High Court’s interim reading in. The Minister of Police partially appeals the judgment and orders of the High Court. More on that later. In the event that the appeal is not upheld, the Minister opposes the confirmation of one part of the order of the High Court. That is the part relating to post-surveillance notification. The Minister of State Security appeals the whole judgment and order of the High Court. The Minister of Justice does not appeal the High Court’s declaration of invalidity, nor does he oppose the application for confirmation. His submissions seek merely to assist this Court. Even though two are appellants, for convenience I collectively refer to the Ministers as the respondents. The remaining respondents have not participated in the proceedings in this Court.
Overview of the legislative framework
 RICA is the result of an overhaul of the Interception and Monitoring Prohibition Act. Its adoption was informed by considerable technological developments in electronic communications, including cellular communications, satellite communications and computer communications. These rendered the old legislation, which mainly revolved around the interception of postal articles and fixed line communications, outdated.
 Focusing on some of the provisions of RICA that are of relevance to this matter, section 2 provides that all forms of interception and monitoring of communications are prohibited unless they take place under one of the recognised exceptions. RICA regulates the interception of both direct and indirect communications, which are defined broadly to include oral conversations, email and mobile phone communications (including data, text and visual images) that are transmitted through a postal service or telecommunication system.
 Without a “designated Judge” RICA would be substantially inoperable. With the exception of only one type, at the centre of all surveillance directions issued under RICA is a designated Judge; she or he must authorise all directions that fall within the purview of functions of a designated Judge. These directions are provided for in sections 16 to 18, 21 and 23. Explaining briefly what each of these sections is about, section 16 provides for “interception directions”. These are directions for the interception of direct or indirect communications.
Section 17 governs the issuing of real-time communication-related directions. In terms of section 18 an application may be made to a designated Judge for combined applications for interception directions, for real-time or archived communication-related directions, or for interception directions supplemented by real-time communication-related directions. Section 21 empowers a designated Judge to issue a decryption direction.
All applications to a designated Judge under sections 16 to 18 and 21 and 22 must be in writing, some even on affidavit. Section 20 authorises the amendment or extension of an existing direction by a designated Judge. Section 22 provides for the issuing by a designated Judge of an entry warrant. Entry in terms of this warrant may be for the purpose of installing an interception device to facilitate interceptions conducted in terms of an interception direction issued under section 16.
 In terms of section 23 applications may be made orally to a designated Judge for the issuing of the directions envisaged in sections 16 to 18 and 21 and an entry warrant envisaged in section 22. An oral application may be made where – because of urgency or exceptional circumstances – it is not reasonably practicable to apply in writing. In terms of section 23(5) directions and entry warrants must be in writing. They too may be oral if that be dictated by urgency or exceptional circumstances.
 Surveillance under these sections, i.e. 16 to 18 and 20 to 23, covers almost the entire spectrum of state surveillance. As shown by this discussion, that wide spectrum involves a designated Judge. What remains is – by comparison – a minuscule aspect provided for in section 19.
 Although directions under section 19 are not issued by a designated Judge, at its centre this section also has a Judicial Officer. This section empowers a High Court Judge, Regional Court Magistrate or Magistrate to issue archived communication related directions. This differs from section 18 in that it applies where only archived communication-related directions are sought. On the other hand, section 18 authorises combined directions where the combinations include archived communication-related directions. And in the case of section 18, the issuer is a designated Judge, not the Judicial Officers mentioned in section 19.
 All surveillance under the various forms of directions is in relation to serious offences; actual or potential threats to the public health or safety, national security or compelling national economic interests of the Republic; organised crime or terrorism; or property which is or could be an instrumentality of serious offences or the proceeds of unlawful activities.
 The facts are not central to the issues before us, but illustrate the potential constitutional difficulties that may arise in the application or misapplication of RICA. In the High Court Mr Sole recounted his undisputed first-hand experience of the abuse of RICA by state authorities. In 2008 he suspected that his communications were being monitored and intercepted. In 2009 he took steps to obtain full disclosure of the details relating to the monitoring and interception of his communications from the Office of the Inspector-General of Intelligence. These efforts were fruitless because – as he was told in a letter – the Inspector-General had found the National Intelligence Agency (NIA) and the crime intelligence division of the police not to be guilty of any wrongdoing. The letter continued that, as RICA prohibits disclosure of information relating to surveillance, Mr Sole could not be furnished with the information. Mr Sole was thus left in the dark as to whether his communications had in fact been intercepted and, if so, what the basis for interception was.
. . . . .
Constitutionality of RICA
 The interception and surveillance of an individual’s communications under RICA is performed clandestinely. By nature, human beings are wont – in their private communications – to share their innermost hearts’ desires or personal confidences, to speak or write when under different circumstances they would never dare do so, to bare themselves on what they truly think or believe. And they do all this in the belief that the only hearers of what they are saying or the only readers of what they have written are those they are communicating with. It is that belief that gives them a sense of comfort – a sense of comfort either to communicate at all; to share confidences of a certain nature or to communicate in a particular manner. Imagine how an individual in that situation would feel if she or he were to know that throughout those intimate communications someone was listening in or reading them.
 If there ever was a highly and disturbingly invasive violation of privacy, this is it. It is violative of an individual’s inner sanctum. In Hyundai Langa DP held that “privacy is a right which becomes more intense the closer it moves to the intimate personal sphere of the life of human beings and less intense as it moves away from that core”. What I have typified – insofar as it relates to the sharing of intimate personal confidences – certainly falls within the “intimate personal sphere”. RICA allows interception of all communications. The sanctioned interception does not discriminate between intimate personal communications and communications, the disclosure of which would not bother those communicating. Nor does it differentiate between information that is relevant to the purpose of the interception and that which is not. In other words, privacy is breached along the entire length and breadth of the “continuum”. And this intrusion applies equally to third parties who are not themselves subjects of surveillance but happen to communicate with the subject. That means communications of any person in contact with the subject of surveillance – even children – will necessarily be intercepted.
 There can be no question that the surveillance of private communications limits the right to privacy. Unsurprisingly, the respondents do not dispute this. Is that limitation reasonable and justifiable under section 36(1) of the Constitution?
Nature of the right
 The country’s apartheid history was characterised by the wanton invasion of the privacy of people by the state through searches and seizures, the interception of their communications and generally by spying on them in all manner of forms. Here is what Mistry tells us:
“The existence of safeguards to regulate the way in which state officials may enter the private domains of ordinary citizens is one of the features that distinguish a constitutional democracy from a police state. South African experience has been notoriously mixed in this regard. On the one hand there has been an admirable history of strong statutory controls over the powers of the police to search and seize. On the other, when it came to racially discriminatory laws and security legislation, vast and often unrestricted discretionary powers were conferred on officials and police. Generations of systematised and egregious violations of personal privacy established norms of disrespect for citizens that seeped generally into the public administration and promoted amongst a great many officials habits and practices inconsistent with the standards of conduct now required by the Bill of Rights. [The right to privacy] accordingly requires us to repudiate the past practices that were repugnant to the new constitutional values, while at the same time re-affirming and building on those that were consistent with these values.”
 Although the focus of this is on searches and seizures, it is very much relevant to the interception of communications. The constitutionally protected right to privacy seeks to be one of the guarantees that South Africa will not again act like the police state that it was under apartheid. Axiomatically, therefore, the right to privacy is singularly important in South Africa’s constitutional democracy.
 To this, one may add the fact that the invasion of an individual’s privacy infringes the individual’s cognate right to dignity, a right so important that it permeates virtually all other fundamental rights. About its importance, Ackermann J said “the right to dignity is a cornerstone of our Constitution”. And in Hugo this Court quoted the words of L’Heureux-Dube J with approval. They are that “inherent human dignity is at the heart of individual rights in a free and democratic society”.
Importance of the purpose of the limitation
 The respondents submit that, notwithstanding the magnitude of the incursion into privacy, the purpose and importance of state surveillance render surveillance under RICA reasonable and justifiable. Its purpose is to investigate and combat serious crime, guarantee national security, maintain public order and thereby ensure the safety of the Republic and its people. In this regard, the Minister of Police explains that the interception of communications for this purpose is part of the fulfilment of the South African Police Service’s obligation under section 205(3) of the Constitution. The other Ministers echo this. The Minister of Justice adds that South Africa is plagued by serious and violent crime which necessitates the adoption of measures such as RICA to detect, investigate and curb serious crimes.
 Without question, it is crucial for the state to secure the nation, ensure that the public is safe and prevent serious crime. These are constitutional obligations. Through RICA, interceptions of communications have come to be central to the fulfilment of these obligations. Thus they serve an important purpose. Unsurprisingly, a number of constitutional democracies have adopted similar measures. This Court has acknowledged that “the rate of crime in South Africa is unacceptably high” and that “the need to fight crime is thus an important objective in our society”. This is as true today as it was when it was said in the year 2000.
Nature and extent of the limitation
 The indiscriminate tentacles of interceptions reach communications of whatever nature, including the most private and intimate. Some of the communications do not in the least have anything to do with the reason for the surveillance. And some of those communicating with the subject of surveillance are collateral victims. I cannot but conclude that the limitation of the right is egregiously intrusive.
 Whilst I must accept that RICA serves an important government purpose, the question to answer here is: is it doing enough to reduce the risk of unnecessary intrusions? In Van der Merwe, this Court considered whether search and seizure warrants were valid despite their failure to mention the offences to which the search related. It held:
“Warrants issued in terms of section 21 of the [Criminal Procedure Act] are important weapons designed to help the police to carry out efficiently their constitutional mandate of, amongst others, preventing, combating and investigating crime. In the course of employing this tool, they inevitably interfere with the equally important constitutional rights of individuals who are targeted by these warrants.
Safeguards are therefore necessary to ameliorate the effect of this interference. This they do by limiting the extent to which rights are impaired. That limitation may in turn be achieved by specifying a procedure for the issuing of warrants and by reducing the potential for abuse in their execution. Safeguards also ensure that the power to issue and execute warrants is exercised within the confines of the authorising legislation and the Constitution.” (Emphasis added.)
 This statement of the law is about safeguards on the exercise of the power of search and seizure in accordance with the Constitution and empowering legislation. At issue here is whether RICA itself does have safeguards that help ensure that the interception of communications and surveillance generally are within constitutionally compliant limits. Put differently, are there safeguards that acceptably minimise the trampling of the privacy right? Let me reiterate what this Court said in a different context in Mistry:
“The existence of safeguards to regulate the way in which state officials may enter the private domains of ordinary citizens is one of the features that distinguish a constitutional democracy from a police state.”
 Subject to RICA, no interceptions of private communications may be effected. Chapter 9 of RICA criminalises interceptions that are at variance with its provisions. RICA provides for the interception of communications in certain instances. Most pertinently, state agents may apply to a designated Judge for authorisation to intercept the private communications of any person, provided certain requirements are met. In summary, RICA requires the application to include the identity of the intended subject, the intended period of interception, the full particulars of the facts and circumstances in support of the application (including details of other investigative procedures that have been applied and failed), a description of the place of interception and type of communications to be intercepted, and the ground on which the application is made. The latter requirement is important; an interception direction may only be issued on certain limited grounds, and – as discussed above – all are of some magnitude.
 However, RICA is silent on a number of crucial issues, many of which have been highlighted by the applicants.
 The United Nations Human Rights Committee’s 2015 report on RICA supports the contention that the structures, processes and safeguards in RICA are insufficient to protect the privacy rights of subjects. To this end, the report states:
“The Committee is concerned about the relatively low threshold for conducting surveillance in the State party and the relatively weak safeguards, oversight and remedies against unlawful interference with the right to privacy contained in [RICA].”
 I turn to specific constitutional challenges on the basis of which the applicants claim the safeguards contained in RICA are inadequate. In this judgment my use of “adequate” or “sufficient” and their derivatives in the context of “safeguards” is not meant to introduce a standard that is higher than the reasonableness and justifiability standard set by section 36(1) of the Constitution. All I mean is: are the safeguards sufficient to meet that reasonableness and justifiability standard?
 We have before us evidence of abuse of the power of surveillance. One example is that of two journalists, Mr Mzilikazi wa Afrika and Mr Stephen Hofstatter, whose phones were tapped by the police’s Crime Intelligence Division while the two were investigating corruption scandals in the South African Police Service. In order to obtain interception directions under RICA, the police told the designated Judge that the phone numbers to be tapped were those of suspected ATM bombers. That, of course, was a lie. The interception direction was granted on the basis of that lie. It authorised the real-time interception of the calls, text messages and metadata of the journalists. These facts have not been disputed by the respondents.
 Regarding the next example, it is immaterial whether actual surveillance had been conducted and – if it was – what the results were. What I am about to narrate manifests the power that state agencies entrusted with the mandate of surveillance wield and the lengths to which they can and do go for purposes best known to them, all facilitated by the fact that they operate in complete secrecy. That is the importance and relevance of this example: the fact of being able to do anything purportedly under RICA. The agencies can obtain interception directions by unlawful means or they may not conduct surveillance at all but produce a negative fictional “intelligence report” about an individual.
They can produce this kind of report even where they have conducted surveillance that did not yield the desired results. Now this next example. In 2006 the Inspector-General concluded a report on surveillance that had been conducted by NIA operatives on prominent South African businessman, Mr Sakumzi Macozoma. The Inspector-General looked into the matter pursuant to a request from the Minister for Intelligence Services in terms of section 7(7)(c) of the Intelligence Services Oversight Act.
The principal motivation of the NIA for the surveillance was stated to be Mr Macozoma’s links with a foreign intelligence service, which were inimical to national security. The Inspector-General’s report found that certain emails had allegedly been intercepted. Purportedly, those emails revealed various conspiracies, which, in turn, allegedly resulted in the electronic and physical surveillance of certain individuals and political parties. The report concluded that the emails were fabricated by the NIA team.
 It would be naïve to think that these examples are odd ones out and that in all other instances state agencies responsible for surveillance have always acted lawfully. The fact that it is now said that the document on the basis of which Mr Sole was subjected to surveillance cannot be found is quite curious; I deliberately put it no higher.
 The last two examples show us that blatant mendacity may be the basis of an approach to the designated Judge. And a designated Judge has no means meaningfully to verify the information placed before her or him. As a result, she or he is left none the wiser. Also, by its very nature – in particular because it takes place in complete secrecy, on the understanding that the subject of surveillance who is best placed to identify an abuse will never know – surveillance under RICA is susceptible to abuse. A key factor which likely emboldens those who conduct surveillance to abuse the process is thus a sense of impunity. The question then is whether lesser restriction on secrecy in the form of notification would thwart the realisation of what RICA interceptions are meant to achieve. Obviously, pre-interception disclosure would defeat the very purpose of surveillance. What about post-surveillance notification?
 The Minister of Police, who appeals the High Court’s order on this issue, is arguing for the retention of the blanket non-availability of notification. He contends that the Constitution confers no right to notification; not pre- or post-surveillance. That, of course, is a misconceived approach. The question is whether – for purposes of the proportionality analysis – denying post-surveillance notification is not overbroad.
 Sections 42(1) and 51 of RICA respectively prohibit and criminalise the disclosure of the fact that an interception direction was issued. Unlike search and seizure warrants which, although also obtained without the knowledge of the subject, do come to the notice of the subject, interception directions under RICA are applied for, granted and implemented in complete secrecy.
Even if a direction ought not to have been granted, all things being equal, the subject will never know. Of course, there cannot be a challenge to the lawfulness of something the subject of surveillance is not aware of. And it is purely fortuitous that some subjects of surveillance do become aware of their surveillance. In the vast majority of cases they never do. That must surely incentivise or facilitate the abuse which we know does take place.
 Therefore, an individual whose privacy has been violated in the most intrusive, egregious and unconstitutional manner never becomes aware of this and is thus denied an opportunity to seek legal redress for the violation of her or his right to privacy. In her or his case the right guaranteed by section 38 of the Constitution to approach a court to seek appropriate relief for the infringement of the right to privacy is illusory. This, at a stage that is post-surveillance and when no prejudice can be suffered by the state agency responsible for the surveillance. That is overbroad and does not help advance the achievement of the purpose of surveillance. What it does instead is to facilitate the abuse of the process under the cloak of secrecy.
 Because of its likely outcomes, post-surveillance notification will go a long way towards eradicating the sense of impunity which certainly exists. The concomitant will be a reduction in the numbers of unmeritorious intrusions into the privacy of individuals. I explain this presently. In a sense, post-surveillance notification functions as less restrictive – or should I say less intrusive – means and serves at least two purposes.
- First, the subject of surveillance is afforded an opportunity to assess whether the interception direction was applied for and issued in accordance with the Constitution and RICA. If need be, she or he may seek an effective remedy for the unlawful violation of privacy.
- Second, because there will be challenges to illegally sought and obtained interception directions, that will help disincentivise abuse of the process and reduce violations of the privacy of individuals.
I am not addressing myself to the possibility that surveillance may, in any event, take place outside of the law, i.e. not even under colour of compliance with RICA. That issue is not before us.
 While internationally there is no consensus on when and how post-surveillance notification is an absolutely necessary safeguard of the right to privacy, considerable comparative practice supports the conclusion that some form of notice is crucial to minimising abuse. Subject to varying exceptions and qualifications, the default position in the United States of America and Canada is to give notice 90 days after surveillance, if safe to do so. Likewise in Denmark, there is a general requirement to inform the subject at the end of surveillance, provided that such notification does not undermine the purpose of the investigation. Germany adopts a similar approach.
 The jurisprudence of the European Court of Human Rights similarly links notification to whether it no longer jeopardises the purpose of the surveillance. This is captured in this passage from Association for European Integration and Human Rights:
“According to the Court’s case law, the fact that persons concerned by such measures are not apprised of them while the surveillance is in progress or even after it has ceased cannot by itself warrant the conclusion that the interference was not justified under the terms of paragraph 2 of Article 8, as it is the very unawareness of the surveillance which ensures its efficacy. However, as soon as notification can be made without jeopardising the purpose of the surveillance after its termination, information should be provided to the persons concerned”. (Emphasis added.)
 Reverting to South Africa, none of the respondents proffered any cogent reason why there should never be post-surveillance notification. And they could not have been able to. I say so because I just cannot conceive of any legitimate reason why the state would want to keep the fact of past surveillance a secret in perpetuity. “[A]s soon as notification can be made without jeopardising the purpose of the surveillance after its termination”, just why would the state want to keep the fact of surveillance secret?
This, of course, is not about those instances where – for a while perhaps – the state may be able to justify why it would be injurious to its interest prematurely to give notification. I am thus led to the conclusion that post-surveillance notification should be the default position, which should be departed from only where, on the facts of that case, the state organ persuades the designated Judge that such departure is justified. And RICA is unconstitutional to the extent that it fails to provide for it. This conclusion is buttressed by the fact that at present the infringement is such as to implicate the rights of access to courts and to an appropriate remedy.
 Plainly, RICA applies to all citizens; the influential and the barely noticed, the well-resourced and the deprived, the well-placed in society and the marginalised, those who can stand and fight for themselves and those who – because of all manner of deprivation – are susceptible to abuse. In South Africa, the vast majority of people cannot afford to litigate where they have suffered the infringement of their rights at the hands of the state. For many, therefore, post-surveillance notification will not translate to the vindication of their privacy rights through the exercise of the right of access to court.
As a result of financial want, exercising the right will be an impossibility. It seems to me what could give this vulnerable group of South Africans a fair chance of also being in a position to vindicate their privacy right would be if they were to be afforded relatively inexpensive, speedy and effective access to judicial review. This could, for example, be in the form of automatic review by the designated Judge in an informal, mainly paper-based non-court process. It could, of course, be open to the designated Judge to call for whatever information she or he might require from whomsoever. The idea is for a summary, but effective process. The detail on how the process should unfold is best left to Parliament.
 In addition to the vindication of the privacy right, here is what commends automatic review. The knowledge by the relevant law enforcement officers that the vast majority of South Africans are most likely not going to be in a position to challenge their surveillance may serve as an incentive for continued abuse. And – as shown by the examples I gave above – abuse is a shocking reality. On the other hand, constant awareness that all interceptions of communications in respect of which RICA has been invoked will be subjected to automatic review will likely serve as a disincentive.
 Automatic review is not an unknown quantity in our legal system. Two examples are the automatic review by Judges of certain sentences imposed by Magistrates and the automatic review by the Land Claims Court of orders of eviction granted in the Magistrates’ Courts. Plainly, these examples are an attempt at guaranteeing justice, and the vast majority of those who stand to benefit are the vulnerable and financially deprived. What we have before us in this application is comparable.
 Automatic review is thus another possible safeguard which, though not highlighted by the applicants, Parliament may consider, given the context in which the current surveillance regime exists. However, it is worth noting that the lack of such a process does not alone render RICA unconstitutional, as the lack of post-surveillance notification does. Automatic review is a complementary mechanism tied to notification.
 To the extent that automatic review would add to the volume of the designated Judge’s work, Parliament is best placed to assess whether the answer would lie in an increase in the number of designated Judges. And the number could be stipulated by Parliament itself or that could be left for the determination of the Minister from time to time. Provision could even be made for the designated Judge initially to enquire from the person concerned if they opt for automatic review or – if so minded – for challenging the surveillance in court.
 Two things are worth noting.
- First, what is said here about automatic review is not meant to serve as precedent for automatic review in all areas where infringements of rights by the state occur.
- Second and linked to the first point, this is but a possible safeguard that Parliament may consider.
It is not obligatory that it be adopted. What is obligatory is for Parliament to put in place a communication surveillance system that sufficiently safeguards against infringements of the privacy right. If – upon evaluation – a new system does serve this purpose even though it does not provide for automatic review, that will pass muster. Initially it lies with Parliament to craft that system. But, of course, it is the judicial system that is the final arbiter on whether the system is constitutionally compliant.
Independence issue and designated Judge
 At the beginning of oral argument, counsel for the applicants was asked to clarify if the independence issue was a standalone challenge or was a facet of the Bill of Rights challenge under section 14, i.e. the privacy challenge. Counsel clarified that the independence issue was a facet of the privacy challenge and that the point being made was that the mooted lack of independence detracted from the sufficiency of safeguards for purposes of the section 36(1) justification exercise. It is on that basis that I deal with the independence issue.
 As shown above, RICA’s framework for surveillance has as its centrepiece a “designated Judge” who authorises surveillance both in real time and of archived communications. Safeguards on the appointment, term, and function of a designated Judge are accordingly pivotal in assessing whether RICA meets the section 36 threshold. I will set out the grounds of the independence challenge later. Before dealing with this challenge, there is an issue that must be resolved first. I next focus on it.
 This issue was not raised by any of the parties. It is that beyond the definition of “designated Judge”, nothing in the rest of RICA provides for the designation or appointment. What arises from this is whether the Minister does have the power to designate a Judge. This question was raised by this Court during argument and debated with counsel. In their affidavits and submissions – written and oral – the parties proceeded from an assumption that the Minister does have the power. They were subsequently invited by way of post-hearing directions to address this issue and its implications in supplementary written submissions. The applicants, the Minister of Justice and Correctional Services and Minister of Police did. All three submit that there is an implied power to designate a Judge.
 Since this issue was not raised by the parties, can this Court consider it mero motu (of its own accord)? This Court in Director of Public Prosecutions, Transvaal held that a court may raise a constitutional issue of its own accord. Additionally, “[w]here a point of law is apparent on the papers, but the common approach of the parties proceeds on a wrong perception of what the law is, a court is not only entitled, but is in fact also obliged, mero motu, to raise the point of law”. This is rooted in the supremacy of the Constitution. It may turn out that the assumption from which the parties proceeded is not wrong. That notwithstanding, in a given case circumstances may be such that a court must raise and determine the legal point underlying the assumption. Since courts are ordinarily required to decide only issues properly raised, constitutional issues should only be raised mero motu in exceptional circumstances:
“The first is where it is necessary for the purpose of disposing of the case before it, and the second is where it is otherwise necessary in the interests of justice to do so. It will be necessary for a court to raise a constitutional issue where the case cannot be disposed of without the constitutional issue being decided. And it will ordinarily be in the interests of justice for a court to raise, of its own accord, a constitutional issue where there are compelling reasons that this should be done.
. . .
It is neither necessary nor desirable to catalogue circumstances in which it would be in the interests of justice for a court to raise, of its own accord, a constitutional issue. This is so because this depends upon the facts and circumstances of a case.”
 The question whether RICA empowers the Minister to designate a Judge falls within both circumstances.
First, this Court cannot dispose of the independence issue without considering the nature and manner of designation. And it can do that by analysing the power of designation and all that pertains to its exercise. To do that, it must know everything that pertains to the designation, for example:
- the exact nature of the powers and functions of a designated Judge;
- the duration of the designation; how that duration comes to an end and the Minister’s role, if any, in that regard; and
- whether, and to what extent, the designated Judge’s powers are subject to the Minister’s control.
It thus becomes necessary for purposes of properly determining this application to deal with the question of the Minister’s power to designate.
 Second, considering the centrality of the designated Judge in the application of RICA and in the matrix of statutes regarding surveillance, the lack of an empowering provision in RICA, if there be, is potentially catastrophic. It is thus also in the interests of justice for this issue to be raised mero motu.
 The question whether RICA empowers the Minister to designate a Judge “is apparent on the papers”. In the face of the challenge that the designated Judge lacks independence because of, inter alia, the manner of designation and concerns around the Judge’s term, one cannot avoid asking: where is the designation provided for and what does it stipulate? For that reason, the question of the existence of the power is a perfect candidate for being raised by the Court mero motu on the papers before us. And I cannot conceive of any prejudice that may be suffered by any of the parties. This is especially so here as the parties were afforded an opportunity to address us on the issue.
 In terms of section 1 of RICA “‘designated Judge’ means any Judge of a High Court discharged from active service under section 3(2) of the Judges’ Remuneration and Conditions of Employment Act [47 of 2001], or any retired Judge, who is designated by the Minister to perform the functions of a designated Judge for purposes of [RICA]”. This definition indicates that RICA meant to empower the Minister to designate a Judge. That this is so is strengthened by the reference in other legislation to the “Judge designated by the Minister of Justice for the purposes of [RICA]”. However, there is only mention of designation in the section 1 definition and nothing in the substantive provisions of RICA expressly empowers the Minister to designate a Judge. Is this power of designation implied in RICA?
 This case presents us with an opportunity to deal not with the common and oft dealt-with necessary or ancillary implied power (which I will simply call the ancillary implied power), but with what I would call a primary implied power. A distinction must be drawn between an implied primary power and an ancillary implied power. I consider it necessary to draw this distinction because quite often discussions of implied powers entail ancillary implied powers, and not primary implied powers.
The distinction will be better understood if I first discuss the well-known concept, the ancillary implied power. An ancillary implied power arises where a primary power – whether express or implied – conferred by an Act cannot be exercised if the ancillary implied power does not also exist.
For example, in Masetlha Moseneke DCJ, considering the President’s power to dismiss a head of an intelligence agency under section 209(2) of the Constitution, held:
“The power to dismiss is necessary in order to exercise the power to appoint. . . . Without the competence to dismiss, the President would not be able to remove the head of the Agency without his or her consent before the end of the term of office, whatever the circumstances might be. That would indeed lead to an absurdity and severely undermine the constitutional pursuit of the security of this country and its people. That is why the power to dismiss is an essential corollary of the power to appoint. . .”
 There, the power to dismiss was found to be an essential corollary of the power to appoint, and this Court thus interpreted the power in section 209(2) of the Constitution to appoint the head of the NIA to include a power to dismiss. The power to dismiss was an ancillary implied power, ancillary because it flowed from the power to appoint. In Matatiele Municipality Ngcobo J wrote:
“It was . . . inevitable that the alteration of provincial boundaries would impact on municipal boundaries. This is implicit in the power to alter provincial boundaries. It is trite that the power to do that which is expressly authorised includes the power to do that which is necessary to give effect to the power expressly given. The power of Parliament to redraw provincial boundaries therefore includes the power that is reasonably necessary for the exercise of its power to alter provincial boundaries.”
 What I refer to as an ancillary power arises in the context of one power being necessary in order for an unquestionably existing power to be exercised.
 Examples of implied powers that I have picked up from academic writings have also been about implied ancillary powers. Hoexter says:
“As a general rule, express powers are needed for the actions and decisions of administrators. Implied powers may, however, be ancillary to the express powers, or exist either as a necessary or reasonable consequence of the express powers. Thus ‘what is reasonably incidental to the proper carrying out of an authorised act must be considered as impliedly authorised’.”
 According to De Ville—
“[w]hen powers are granted to a public authority, those granted expressly are not the only powers such public authority will have. The powers will include those which are reasonably necessary or required to give effect to and which are reasonably or properly ancillary or incidental to the express powers that are granted.”
 Baxter says something that is to similar effect as what Hoexter and De Ville say. He says:
“Powers may be presumed to have been impliedly conferred because they constitute a logical or necessary consequence of the powers which have been expressly conferred, because they are reasonably required in order to exercise the powers expressly conferred, or because they are ancillary or incidental to those expressly conferred.”
 Coming to an implied primary power, an antecedent question is: what do I mean by a primary power? A primary power is a power to do something required to be done in terms of an Act and which does not owe its existence to, or whose existence is not pegged on, some other power; it exists all on its own. That is what makes it primary, and not ancillary. If it owed its existence to another primary power, then it would be an ancillary power.
 A primary power may be express or implied. It is express if it is specifically provided for. Examples of express primary powers are the President’s power to appoint the head of the intelligence agency, which featured in Masetlha, and the power to determine or alter provincial boundaries, which featured in Matatiele Municipality. The primary power is implied if it is not expressly provided for. It is implied from a reading of the Act and a consideration of all that must be factored in the interpretative exercise. It owes its existence to provisions of the Act and everything that is relevant to the interpretative exercise. The fact that provisions of the Act, including provisions conferring other primary powers, may shed light on whether an implied primary power exists does not mean the implied primary power derives its existence from these provisions. These provisions and all that must be factored in determining whether a primary implied power exists serve as interpretative tools that point to its existence. As we now know, the Constitution plays a crucial role in that interpretative exercise.
 So, the interpretative exercise is not confined to the four corners of a statute. The answer to the question whether an implied primary power exists is yielded by the usual interpretative exercise that seeks to establish what a statute or a provision in it means. There is nothing unusual about this. That, in the main, available jurisprudence and learning on implied powers concerns ancillary implied powers does not mean there aren’t primary implied powers. And cases like Masetlha and Matatiele Municipality never suggested otherwise. If we were to say they did, we would be erring. They pronounced on something different, which is what was before them. That is, in each instance the question was whether the ancillary implied power at issue existed. They did not in the least purport to declare that there is nothing like an implied primary power. And they could not have so declared as that was not before the Court. In other words, they did not suggest that an interpretative exercise can never yield the result that an implied primary power is provided for in a statute. To put it bluntly, the question whether there may be an implied primary power was irrelevant to the issues at hand in those cases. Thus those cases are of no assistance to us in the present matter.
 What I must answer next is whether the power to designate a Judge does exist. What is clear is that if it does, it is not express, nor is it an ancillary implied power. It is not a cognate implied power pegged on, and owing its existence to, some primary power.
 For present purposes, the Promotion of Administrative Justice Act (PAJA) is a useful guide. Section 1 of PAJA provides an example that similarly talks about “designation” by the Minister. A ringingly similar feature is that the designation by the Minister is also in the definition section and the substantive provisions of PAJA say nothing more about the conferral of a power to designate.
Section 1 of PAJA defines “court” to include “a Magistrate’s Court . . . designated by the Minister by notice in the Gazette”. One may be led to believe that this definition is merely descriptive and points to the Minister being empowered elsewhere in PAJA to designate Magistrates’ Courts. But nowhere in PAJA is there such empowering provision. Section 9A of PAJA twice refers to a “Magistrate’s Court designated by the Minister in terms of section 1 of this Act”.
A reading of section 9A and the definition section put it beyond question that the power to designate is conferred by the definition section itself. Plainly, PAJA meant for the power to designate to stem directly from the definition. Put differently, the empowerment of the Minister to designate is implied in the definition in section 1. And this power to designate is a primary power implied from the language and context of PAJA; it is not ancillary to another power in PAJA.
 It seems to me it would be unduly formalistic to suggest that – in the absence of a substantive provision in PAJA providing for the power to designate a Magistrate’s Court – PAJA has not made provision for the power to designate. It is plain from the definition that it is envisaged that designations may be made by the Minister. It also seems clear that the Minister need do nothing more than to reflect the court she or he has designated in the Gazette. It would be too formalistic to expect that there be a substantive provision saying the Minister has a power to designate a Magistrate’s Court by notice in the Gazette. What would that substantive provision add? Nothing at all.
 Of importance, what is to be gleaned from the definition of “court” in section 1 of PAJA is that the conferral of a power by a provision in the definition section is not something unknown. So, we must approach the definition of “designated Judge” in RICA with this awareness in mind and be wary not instinctively to reject the notion that a definition section may well be a power-conferring provision. Of course, RICA is a different statute and it must be its own interpretation that yields a determination that the power to designate exists or does not. My reliance on section 1 of PAJA does not, in the least, mean I am not alive to this. What I seek to point out is that it is not unheard of that a power to do something may be implied from the definition section.
 Reverting to RICA, the definition of “designated Judge” refers to a Judge falling within one of the two identified categories and “who is designated by the Minister to perform the functions of a designated Judge for purposes of [RICA]”. The definition tells us what the functions of the designated Judge are; they are to be found in RICA. And they have largely been identified above.
So, we know who qualifies to be a designated Judge. We know what a designated Judge is required to do. What remains is the formal act of designating, which – in essence – need not entail more than to identify a Judge falling within the two identified categories and advising her or him that she or he has been so identified “to perform the functions of a designated Judge for purposes of [RICA]”.
Why then can’t the Minister designate? Reading the definition and the provisions on the functions of a designated Judge together and taking into account the fact that the act of designating need not go beyond what I have just identified, it would be the height of formalism to insist that the power to designate must be expressly provided in the substantive provisions of RICA.
Yes, a substantive provision conferring the power would have been a “nice-to-have”, but I do not agree that its absence must legally result in a lack of power.
Crucially, the many provisions on the functions of a designated Judge appear to proceed from the premise that the power to designate a Judge does exist in RICA. These provisions must have been meant to be operable based on the definition of “designated Judge” in section 1. To suggest otherwise would be to place form ahead of substance.
The conclusion has to be that the power to designate a Judge is implicit in a proper conjoined reading of the definition of “designated Judge” and other provisions of RICA.
 Also, in accordance with the maxim ut res magis valeat quam pereat, rather than render RICA virtually inoperable as a result of a perceived lack of power to designate, an interpretation that finds a power to designate a Judge in section 1, read with the other provisions I have referred to, commends itself. The maxim may be a useful tool of interpretation. And it is here.
In Hess the Court held that, in terms of this maxim, “[w]here the meaning of a section in a law is uncertain or ambiguous it is the duty of the Court to consider the law as a whole, and compare the various sections with each other and with the preamble, and give such meaning to the particular section under consideration that it may, if possible, have force and effect”. The reading I am advocating is quite viable. Hoexter – albeit in the context of necessary ancillary powers – argues that “[t]here is a very strong argument in favour of implying a power if the main purpose of the statute cannot be achieved without it”.
 If there is no power to designate, no Judge can be designated lawfully. On the Fedsure and Pharmaceutical Manufacturers principle, it would be unconstitutional for the Minister to designate. As the role of the designated Judge is key to RICA surveillance, the lack of the power to designate hollows the Act out and leaves it bereft of meaningful operability. Take the designated Judge out of the picture, directions under sections 16 to 18 and 21 and 23, extensions and amendments of these directions under section 20 and entry warrants under section 22 cannot be issued.
The only directions that can still be issued are archived communication-related directions under section 19. And those are issued, not by a designated Judge, but by a High Court Judge, a Regional Court Magistrate or a Magistrate. The only surveillance that remains possible where a designated Judge would otherwise have had to grant authorisation is in instances where urgency or exceptional circumstances require of law enforcement officers to act without first seeking authorisation from a designated Judge. Surely those instances must constitute a small percentage of surveillance conducted by law enforcement agencies. So, this does not detract from the fact that without a designated Judge, the RICA edifice becomes substantially inoperable; a far-reaching result.
 Faced with that ominous terminal reality, I can conceive of no compelling reason for not concluding that the power to designate is implied in the definition of “designated Judge” in section 1 of RICA; it is an implied primary power.
Considering section 1 with the structure and purpose of RICA as a whole, this seems the only viable interpretation. The only argument against this that I can think of is purely the lack of express provision in the substantive provisions of RICA conferring the power to designate. Surely, that cannot of necessity be dispositive of the question.
Not when we know from sections 1 and 9A of PAJA, considered with the principles underlying implied powers, that a definition section can confer a power. In sum, I conclude that section 1 of RICA, read with the provisions on the functions of a designated Judge, does provide for the power to designate a Judge.
 I am yet to deal with the independence challenge. One may ask: what is the point of “saving” RICA insofar as the question of the power to designate is concerned, if we may still invalidate it based on the independence challenge? These are two distinct aspects that must be dealt with separately. It would be illogical to say there is no power because, in any event, the power-conferring provision is unconstitutional for inconsistency with the Bill of Rights.
There is an antecedent question. That is, is there a power? If there is, there is a second question: is the power constitutional? I now proceed to deal with the independence challenge.
 In their pleaded case the applicants contend that the definition of “designated Judge” which – as I have now held – is the power-conferring provision does not provide adequate safeguards for the structural and perceived independence of the designated Judge.
The independence challenge is founded on the grounds that:
- RICA fails to prescribe or limit the designated Judge’s term of office, making it possible for the Minister to make indefinite reappointments;
- each term is for a duration determined at the whim of the Minister; and
- appointments of designated Judges are exclusively made by a member of the Executive in a non-transparent manner in that there is no role for the Judicial Service Commission (JSC), Parliament or the Chief Justice.
The Minister of State Security argues that the Constitution is silent on the appointment of Judges to perform the functions contemplated in RICA. She stresses that the designated Judge is appointed from the ranks of Judges who are presumed independent, and that the Constitution does not require extra measures to guarantee their independence.
 That search and seizure warrants – which by their nature result in the violation of privacy – must be issued by an independent Judiciary is a leitmotif across our constitutional jurisprudence.
In Thint this Court recognised “the fact that the decision as to whether a warrant is to be issued is taken by an impartial and independent Judicial Officer . . . as an important consideration in determining the constitutionality of search powers”, and held that “requiring a search warrant to be issued by a Judicial Officer is an important part of the protection of fundamental rights and, in particular, the right to privacy”.
Plainly, that puts it beyond question that it is a constitutional requirement that the issuing of search and seizure warrants be authorised by an independent Judicial Officer.
 This concern for independent judicial authorisation of intrusions into privacy echoes the ruling in Van der Merwe, regarding the validity of search and seizure warrants issued in terms of section 21 of the Criminal Procedure Act.
Mogoeng J stated that—
“[t]he judicious exercise of this power by [Judicial Officers] enhances protection against unnecessary infringement. They possess qualities and skills essential for the proper exercise of this power, like independence and the ability to evaluate relevant information so as to make an informed decision”. (Emphasis added.)
 Why this is crucial is because – by its very nature – the execution of warrants of search and seizure results in the violation of privacy. The involvement of independent functionaries like members of the Judiciary helps ensure that the risk of unmeritorious intrusions into the privacy of individuals is minimised.
This must apply a fortiori (with more force) in the case of surveillance under RICA. That is so because the non transparent, if not impenetrable, circumstances in which the power of issuing RICA surveillance directions is exercised make it singularly important that there be no apprehension or perception of lack of independence; more important than in the case of issuing search and seizure warrants where the possibility of a challenge is always a reality. And the operative words are “apprehension” and “perception”.
 Once satisfied that the Constitution requires an independent designated Judge to authorise interceptions, it is necessary to ascertain the meaning of independence.
The Supreme Court of Canada in R v Valente defined independence thus:
“The word ‘independent’ . . . connotes not merely a state of mind or attitude in the actual exercise of judicial function, but a status or relationship to others, particularly to the Executive branch of government, that rests on objective conditions or guarantees.”
 This Court in McBride, in considering the independence of the Independent Police Investigative Directorate, explained that—
“it is difficult to attempt to define the precise contours of a concept as elastic as [independence]. It requires a careful examination of a wide range of facts to determine this question. Amongst these are the method of appointment, the method of reporting, disciplinary proceedings and method of removal of the Executive Director from office, and security of tenure.” (Emphasis added.)
 In examining the elements of independence, this Court in Glenister II accepted that the question is not whether an institution has absolute independence, but rather “whether it enjoys an adequate level of structural and operational autonomy that is secured through institutional and legal mechanisms designed to ensure that it ‘discharges its responsibilities effectively’, as required by the Constitution”.
Additionally, the perception of independence plays a critical role in ascertaining whether an institution is independent.
In particular, “[w]hether a reasonably informed and reasonable member of the public will have confidence in an entity’s autonomy protecting features is important to determining whether it has the requisite degree of independence”.
 Admittedly, many of this Court’s judgments regarding independence deal with inspectorates and other institutions, and not with the independence of Judges. One case where focus was on judicial officers is Van Rooyen where Chaskalson CJ said:
“In De Lange v Smuts N.O. and Others, Ackermann J referred to the views of the Canadian Supreme Court in The Queen in Right of Canada v Beauregard, Valente v The Queen and R v Généreux on the question of what constitutes an independent and impartial court, describing them as being ‘instructive’. In this context, he mentioned the following summary of the essence of judicial independence given by Dickson CJC in Beauregard’s case:
‘Historically, the generally accepted core of the principle of judicial independence has been the complete liberty of individual judges to hear and decide the cases that come before them; no outsider – be it government, pressure group, individual or even another judge – should interfere in fact, or attempt to interfere, with the way in which a judge conducts his or her case and makes his or her decision. This core continues to be central to the principle of judicial independence.’
This requires judicial officers to act independently and impartially in dealing with cases that come before them, and at an institutional level it requires structures to protect courts and judicial officers against external interference.”
 It is generally expected and accepted that Judges act independently, impartially and fairly, without bias or prejudice. This is supported by the burdensome onus on an applicant for recusal of a Judge, which is based on the presumption that Judicial Officers are impartial.
But it does not mean that Judges are infallible, nor does it mean there is no possibility that Judges may attempt to abuse their office. As Judges, we should not – because of our own belief in our independence – be sanctimoniously dismissive of the fact that we may be susceptible to these ills.
Indeed, Nelson explains that “it would not be realistic” for Judges to “assume a sanctimonious or puritanical attitude”.
 As such, structures, mechanisms and processes that strengthen the independence of an institution or office are imperative even where the incumbent is a Judge. Indeed, in Justice Alliance, this Court held that
“[n]on-renewability [of the term] fosters public confidence in the institution of the Judiciary as a whole, since its members function with neither threat that their terms will not be renewed nor any inducement to seek to secure renewal”.
The fact that it was the prestigious office of the Chief Justice that was at the heart of the enquiry in that matter did not render the scrutinisation of its structural or perceived independence unnecessary. Put differently, the fact that it is generally expected and accepted that Judges act independently, impartially and fairly, without bias or prejudice did not stand in the way of this Court considering the independence challenge in relation to the office of Chief Justice.
 Judges are appointed by the President after a constitutionally prescribed process of consultation which differs according to the judicial office applied for. In terms of section 174(3) of the Constitution, the JSC must also be consulted.
In practical terms, the JSC is able to participate meaningfully in that consultation process by subjecting candidates for appointment as Judges to an interview process. Needless to say, the object of the interviews is to determine whether candidates are suitable for the judicial office applied for. The interview process is rigorous and public.
Acting Judges of the Constitutional Court are appointed by the President on the recommendation of the Minister of Justice, acting with the concurrence of the Chief Justice.
And acting Judges of other courts are appointed by the Minister of Justice after consulting the senior Judge of the relevant court. The criteria for selection of Judges are publicly accessible. And once appointed, Judges’ terms of office are strictly regulated. Court hearings are almost exclusively conducted in open court, and reasons for Judges’ decisions are largely accessible. If dissatisfied with the outcome, parties can generally appeal or seek the review of a decision. These processes allow for public scrutiny, accountability and public trust.
 None of these protective processes and structures are in place for the designated Judge under RICA.
At present, the designated Judge is appointed by the Minister of Justice – a member of the Executive – without the involvement of any other person or entity. And in practice, terms of designated Judges have been renewed. Also, the lack of specificity on the manner of appointment and extensions of terms raises independence concerns.
As Justice Alliance tells us, the power to extend the term of a Judge goes to the core of the tenure of the judicial office, judicial independence and the separation of powers, and that open-ended discretion in respect of appointments may raise a reasonable apprehension or perception that the independence of the Judge may be undermined by external interference by the Executive. It is so that the Minister does not appoint a designated Judge as a Judge, but rather as a designated Judge; the designated Judge is already a Judge.
But the point is that – based on Thint and Van der Merwe – the requirement of independence is a constitutional imperative. And there may be factors that conduce to a perception of lack of independence even in the case of a Judge. That, indeed, was one of the underlying bases even for the Justice Alliance ratio (basis for a decision). The designation by a member of the Executive in ill-defined circumstances or circumstances that completely lack description does not conduce to a reasonable perception of independence.
 Additionally, directions are sought and issued in complete secrecy, and, therefore, without public scrutiny, or the possibility of review or appeal, which points to a lack of “mechanisms for accountability and oversight”. My reference to lack of review has in mind not those instances where the subject of RICA surveillance may fortuitously become aware of her or his surveillance.
All of these considerations indicate the lack of structural, operational and perceived independence of the designated Judge. To exacerbate the situation, this lack of structural independence may also lead to a reasonable perception of lack of independence. Obviously, this is something Parliament may address with relative ease.
 RICA is thus declared unconstitutional to the extent that it fails to ensure adequate safeguards for an independent judicial authorisation of interception.
Ex parte issue
 Section 16(7)(a) of RICA provides that “[a]n application must be considered and an interception direction issued without any notice to the person or customer to whom the application applies and without hearing such person or customer”. The rationale for this ex parte process is obvious: the surveillance would be futile if the subject were to be aware of it. The applicants do not challenge this.
 However, the result is that an application for an interception direction that may severely and irreparably infringe the privacy rights of the subject is granted on the basis of information provided only by the state agency requesting the direction. The designated Judge is required to issue the direction on the basis of that one-sided information.
Save perhaps for relatively obvious shortcomings, inaccuracies or even falsehoods, the designated Judge is not in a position meaningfully to interrogate the information. For that reason, as evidenced by the example of the surveillance of Mr Mzilikazi wa Afrika and Mr Stephen Hofstatter, surveillance directions may be issued on unadulterated lies.
And the designated Judge could not be any the wiser. According to the Minister of State Security, there is no cause for complaint because RICA contains sufficient safeguards to ensure that the designated Judge has all the facts and circumstances justifying the surveillance direction. Of course, it is simply not correct that the designated Judge is in a position to confirm the veracity of every detail furnished to her or him.
 In the High Court the applicants argued that the ex parte process undermines the audi alteram partem (listen to the other side) principle and thus violates the right to a fair hearing guaranteed by section 34 of the Constitution.
The argument continued that this is exacerbated by the fact that the interception direction is final, unlike most other ex parte orders, which are granted on an interim basis.
As a result, the applicants suggested that there should be some form of adversarial process to ensure that the interests of the subject of surveillance are properly protected and ventilated before an interception direction is granted. According to them the task of safeguarding the interests of the subject should be performed by a “public advocate”.
 While the High Court agreed with the applicants that the ex parte nature of the process renders RICA unconstitutional, it held that it was not well-placed to consider the security risks, or the method for selecting, vetting and briefing public advocates. It further noted that there are other options that could be adopted to deal with the ex parte issue.
As a result, it simply declared section 16(7) unconstitutional “to the extent that it fails to provide for a system for a public advocate or other appropriate safeguards to deal with the fact that the orders in question are granted ex parte”. It did not grant any interim relief.
 In this Court, the applicants again refer to the institution of a public advocate as a potential solution to the ex parte issue. However, here they argue that the option merely demonstrates that less restrictive means exist. I agree only that surely means that can temper the effects of the ex parte nature of the process do exist. The risk of abuse of the ex parte process highlights the general deficiencies in RICA.
With more and better safeguards, chances of unlawful intrusions into privacy would be minimised. I prefer not to comment on the participation of a public advocate as one such safeguard. The choice of safeguards to address the inadequacies resulting from the ex parte nature of the process is something best left to Parliament. And those need not be an adversarial process. If Parliament opts for that process, that is its choice.
 In sum, RICA is unconstitutional to the extent that it lacks sufficient safeguards to address the fact that interception directions are sought and obtained ex parte.
Management of information issue
 Once a person’s communications are intercepted and obtained, does RICA contain safeguards relating to how that information is handled, stored, and eventually destroyed? The applicants contend that RICA lacks procedures for examining, copying, sharing, sorting through, using, storing or destroying surveillance data. RICA is thus unconstitutional to the extent of this shortcoming.
The applicants’ concern is that the lack of regulation in this regard exposes subjects of interceptions to even more aggravated intrusions into their privacy. The Minister of State Security argues that section 35(1)(f) and (g), read with section 37 of RICA, provides the necessary procedures and that, in any event, section 42 provides for a general prohibition of the disclosure of intercepted communications. The High Court upheld the applicants’ contention.
 Section 35(1)(f) does no more than to enjoin the Director of the Office for Interception Centres (director) to prescribe what information must be kept in terms of section 37 by the heads of information centres. Such information must include particulars relating to: applications for the issuing of directions; the directions themselves; and the results obtained pursuant to the execution of each direction. As the applicants correctly argued, reference to “particulars” means all that the director need prescribe must be kept is an identification of both the application for directions and the directions issued, and an indication of the nature of the result obtained pursuant to the direction.
That means it is not obligatory for the director to prescribe that the actual application, direction and data unearthed by the surveillance must be kept in terms of section 37. That is so because “particulars of an application”, “particulars of a direction” or “particulars of the result” are not the same as an application, direction or result. If the object of the section was to refer to an application, direction or result, it is oddly convoluted to say “particulars relating to applications, directions or results”.
 To the extent that section 35(1)(f) provides that the information in respect of which the director must prescribe “must include” what the section specifies, that means what is prescribed may be more than that. That does not assist the Minister. It cannot be left to the director to make a choice on so important an issue. That choice effectively leaves it to chance whether a director may prescribe that the information that must be kept must include actual applications, directions and results.
This to me seems to be the most important information that a subject of surveillance would have an interest in. That is the type of information that would help her or him make an informed decision whether to litigate for the vindication of rights. Plainly, the threat of litigation has the potential of limiting abuse and, concomitantly, the intrusion into the privacy of individuals. Now whether this important information will be part of what is prescribed to be kept is left to the unbounded discretion of the director. That simply cannot be. There needs to be clear parameters on the exercise of discretion.
 Section 35(1)(g) provides that the director must prescribe the manner in, and the period for, which the information referred to in section 35(1)(f) must be kept. Section 35(1)(g) is thus of no assistance because it refers us back to section 35(1)(f). So, whatever it says is also tainted by whatever shortcomings section 35(1)(f) has.
 Section 37(1) also refers back to section 35(1)(f). It provides that the head of an interception centre must keep or cause to be kept proper records of such information as may be prescribed by the director in terms of section 35(1)(f). Obviously, this too does not assist the case of the Minister of State Security.
In terms of section 37(2) heads of interception centres must on a quarterly basis or as required by the director submit reports to the director. For present purposes, the reports must deal with the records kept in terms of section 37(1). I need not say what has now become the usual refrain of the inutility of these sections as they all have as their base the deficient section 35(1)(f).
 Section 42, another section relied upon by the Minister of State Security, contains a blanket prohibition on the disclosure of information obtained through the application of RICA. Absent detail on what information must be kept and how it must be kept, this does not help much.
 Looked at in their totality, all the sections invoked by the Minister of State Security to meet the applicants’ challenge fail to address the substance of the complaint. And it is a valid complaint.
The sections give no clarity or detail on:
- what must be stored;
- how and where it must be stored;
- the security of such storage;
- precautions around access to the stored data (who may have access and who may not);
- the purposes for accessing the data; and
- how and at what point the data may or must be destroyed.
Thus there is a real risk that the private information of individuals may land in wrong hands or, even if in the “right” hands, may be used for purposes other than those envisaged in RICA. All this exacerbates the risk of unnecessary intrusions into the privacy of individuals. The minimum standards for the proper management of surveillance data were set out as follows by the European Court of Human Rights (ECHR) in Weber:
“[T]he nature of the offences which may give rise to an interception order; a definition of the categories of people liable to have their telephones tapped; a limit on the duration of telephone tapping; the procedure to be followed for examining, using and storing the data obtained; the precautions to be taken when communicating the data to other parties; and the circumstances in which recordings may or must be erased or the tapes destroyed”. (Emphasis added.)
 In this regard, RICA thus fails to provide for means that guarantee lesser invasions of privacy. As a result, the extent of the limitation of the privacy right becomes more egregious. And there is no relation between the purpose of state surveillance and the absence of procedures to safeguard private intercepted communications.
RICA is, therefore, unconstitutional to the extent that it fails adequately to prescribe procedures to ensure that data obtained pursuant to the interception of communications is managed lawfully and not used or interfered with unlawfully, including prescribing procedures to be followed for examining, copying, sharing, sorting through, using, storing or destroying the data.
 The third amicus, Privacy International, contends that the High Court order is too narrow. It notes that, although the applicants’ challenge before the High Court also related to communication data in the hands of private telecommunication service providers, the High Court order is limited to information intercepted and retained by the state.
It submits that there is no difference – from the perspective of intrusions into privacy – between data retained by the state, on the one hand, and that retained by private entities, on the other. It urges us to amend the order to add that RICA is unconstitutional also to the extent that it fails to prescribe proper procedures to be followed when a telecommunication service provider is examining, copying, sharing, sorting through, using, storing or destroying archived communications-related data.
 Essentially, the third amicus is lodging its own appeal. Ordinarily (and I use this word guardedly), an amicus participates in proceedings to raise “new contentions which may be useful to the Court”.
This same idea was captured by this Court in In re Certain Amicus Curiae Applications where it held that “the special duty” of an amicus to the court “is to provide cogent and helpful submissions that assist the court”. (My emphasis).
A court’s task is to determine the dispute presented to it by the parties. It stands to reason then that assistance to it must relate to the determination of that dispute. Adding a different dispute – like an additional appeal – not litigated by the parties is not assistance with the dispute before the court. If anything, that amounts to burdening the Court with something else to determine. That is not what rule 10 and the In re Certain Amicus Curiae Applications statement of law envisage.
 Therefore, it seems to me that it is not in the interests of justice to entertain the issue raised by the third amicus. That said, I will not categorically decide the question whether an amicus may lodge an appeal outside of the application or appeal being litigated by the parties. That is not necessary here.
I limit what I say to a conclusion that is based on the interests of justice in the circumstances of this case; interests of justice do not dictate that the additional appeal of the third amicus be entertained. It is so that in our age of mass data surveillance, private actors arguably pose a comparable threat to privacy as does the state. So, one cannot make light of the issues raised in the appeal by the third amicus. Although the arguments of the third amicus on the issues are compelling, they were not adequately ventilated by the parties. That is understandable because they do not relate to the issues in dispute between the parties.
Practising lawyers and journalists issue
 Before the High Court the applicants contended that a practising lawyer owes a client a duty to keep communications between them confidential, and that it is essential to the practice of journalism for journalists to preserve the confidentiality of the identities of their sources. The argument continued that the surveillance of journalists constitutes a limitation of the right to freedom of expression and the media under section 16. And the surveillance of a practising lawyer infringes legal professional privilege.
Because this privilege is an essential part of the right to a fair trial, the surveillance of lawyers limits the rights to a fair hearing and trial, respectively guaranteed under sections 34 and 35(3) of the Constitution. That, of course, is in addition to limiting the right to privacy. The applicants then sought an order declaring RICA unconstitutional to the extent that it does not expressly address the circumstances where a subject of surveillance is either a practising lawyer or journalist. The High Court agreed.
 Before us the applicants clarified that invoking sections 16, 34 and 35(3) is not meant to constitute three discrete constitutional challenges. Rather, this must be considered within the scope of the section 36 enquiry. In other words, the argument that RICA’s limitation of the right to privacy is not reasonable and justifiable is buttressed by the interplay between the rights protected by these sections and the privacy right.
 Let me preface whatever I am going to say about this challenge by making this point. To state the obvious, it can never be suggested that practising lawyers and journalists should not be subjected to surveillance, which includes the interception of their communications.
They are not immune to the very same type of conduct that justifies the surveillance of other members of society. There may be reasonable grounds of suspecting them of being guilty of, for example, serious criminality or conduct that places the security of the Republic at serious risk. Indeed, in certain instances they may actually be guilty of such conduct.
 I agree that keeping the identity of journalists’ sources confidential is protected by the rights to freedom of expression and the media. This Court has acknowledged the constitutional importance of the media in our democratic society, and has confirmed that
“[t]he Constitution thus asserts and protects the media in the performance of their obligations to the broader society, principally through the provisions of section 16”.
It follows that the confidentiality of journalists’ sources, which is crucial for the performance by the media of their obligations, is protected by section 16(1)(a). Like the High Court, I place reliance on Tsoka J who held as much in Bosasa. Relying on local and foreign authorities, he put it thus:
“[I]t is apparent that journalists, subject to certain limitations, are not expected to reveal the identity of their sources. If indeed freedom of press is fundamental and sine qua non for democracy, it is essential that in carrying out this public duty for the public good, the identity of their sources should not be revealed, particularly, when the information so revealed, would not have been publicly known. This essential and critical role of the media, which is more pronounced in our nascent democracy founded on openness, where corruption has become cancerous, needs to be fostered rather than denuded.”
 I agree with the applicants that legal professional privilege is an essential part of the rights to a fair trial and fair hearing. This Court in Thint held:
“The right to legal professional privilege is a general rule of our common law which states that communications between a legal advisor and his or her client are protected from disclosure, provided that certain requirements are met. The rationale of this right has changed over time. It is now generally accepted that these communications should be protected in order to facilitate the proper functioning of an adversarial system of justice, because it encourages full and frank disclosure between advisors and clients. This, in turn, promotes fairness in litigation. In the context of criminal proceedings, moreover, the right to have privileged communications with a lawyer protected is necessary to uphold the right to a fair trial in terms of section 35 of the Constitution, and for that reason it is to be taken very seriously indeed.”
 This means that, although originally sourced from the common law, legal professional privilege is now undergirded by the Constitution. The proper functioning of our legal system is reliant on the confidentiality of communications between lawyer and client. That in turn promotes the rule of law. Thus the wholesale interception of lawyer-client communications without any recognition of this legal, indeed constitutional, reality would be at odds with the rule of law.
 This would not be an isolated carve out. Canada’s section 186(2) of the Criminal Code prohibits authorisation of an interception “at the office or residence of a solicitor” without “reasonable grounds to believe that the solicitor” or his associates are about to commit an offence.
And the United Kingdom’s Interception of Communications Code of Practice pursuant to section 71 of the Regulation of Investigatory Powers Act 2000 establishes stringent safeguards where the communications intercepted may be of a confidential nature, including journalistic or legally privileged communications.
 In sum, the confidentiality of lawyer-client communications and journalists’ sources is particularly significant in our constitutional dispensation. There is thus a need that special consideration be given to this fact when interception directions are sought and granted. Plainly there are means that may help minimise this particularly egregious form of intrusion into privacy; particularly egregious because of its impact on other constitutional rights. Some of the foreign examples tell us as much. While reference to them should not be seen as dictation to Parliament, they serve as examples of less restrictive means, which do not subvert the purpose of RICA. RICA is thus unconstitutional to the extent that, when the intended subject of surveillance is a practising lawyer or a journalist, it fails to provide for additional safeguards calculated to minimise the risk of infringement of the confidentiality of practising lawyer and client communications and journalists’ sources.
 I should perhaps make the point that before us we do not have other professions seeking similar relief on the basis that they too are deserving of special protection. Nor have the applicants made out such a case. Therefore, we need not consider that aspect. The first amicus, Media Monitoring Africa Trust, has referred to civil society actors and children. This is being raised for the first time before this Court.
 We do not have the benefit of the views of the High Court on the issue. It is by no means an issue capable of easy resolution. Take, for example, civil society actors. While I understand that some of them perform a watchdog function, I do not readily see the confidentiality of their communications to have as strong a constitutional claim to special treatment as does that of journalists and practising lawyers. I am not saying they do not have that claim; it is just not readily apparent to me. But that is only a prima facie view.
Also, how exactly in a judgment like this would we describe “civil society actors” for Parliament to know what we are talking about? Maybe if we try hard enough, we can. But must we? I think not. The term is rather nebulous for us to stray into that terrain as a court of first and last instance on the issue. Also, for reasons similar to those I expressed in respect of the additional appeal by the third amicus, it is doubtful that this relief is properly before us. I need say nothing more.
 The interception of children’s communications is on a different footing. Section 28(2) of the Constitution demands the paramountcy of children’s best interests in all matters affecting them. This Court in Centre for Child Law held that the “analysis of the right to privacy is even more pressing when dealing with children”. Is that enough for us to grant the relief sought by the first amicus?
Most people have contact and communicate with children in one form or another. It is likely that a substantial number of intercepted communications will include communication by or with children. This is an issue that affects children who are the subjects of surveillance and those who are not. Must a special carve out be made only for the second category of children, or for both categories?
If for both categories, must there be a distinction that gives recognition to the fact that children in the first category are the subjects of an interception, and, if so, what should the nature of the distinction be? These possibilities, if not imponderables, underscore the fact that this is just too unfirm a terrain for us to venture into under the circumstances in which the issue has been raised before us.
 To summarise, it is not in the interests of justice to decide the issues concerning communications of children and civil society actors. We would have benefited from their ventilation before, and determination by, the High Court.
Bulk communications surveillance
 According to the respondents, bulk surveillance is—
“an internationally accepted method of strategically monitoring transnational signals, in order to screen them for certain cue words or key phrases. The national security objective is to ensure that the State is secured against transnational threats. It is . . . done through the tapping and recording of transnational signals, including, in some cases, undersea fibre optic cables.”
It may also be explained to relate to
“intelligence obtained from the interception of electromagnetic, acoustic and other signals, including the equipment that produces such signals. It also includes any communication that emanates from outside the borders of [South Africa] and passes through or ends in [South Africa].”
Bulk surveillance thus involves, inter alia, the interception of all internet traffic that enters or leaves South Africa, including the most personal information such as emails, video calls, location, and browsing history.
 The question before us is this: is there a legal basis for the state to conduct bulk communications surveillance? Before answering this question, we must determine whether the appeal in relation thereto is properly before us.
Appeal by Minister of State Security
 The Minister of State Security has filed a notice of appeal that was not accompanied by an application for leave to appeal directly to this Court in respect of the order relating to bulk surveillance. For that reason, the applicants argue that this part of the appeal is fatally defective. That is so because – continues the argument – the High Court order relating to this part of the appeal does not involve any declaration of a statute invalid.
Rather, what the High Court did was to declare unlawful bulk surveillance conducted by state agencies in the absence of a law authorising it. It is only in respect of the issues on which the applicants are seeking confirmation that the Minister of State Security had an automatic right of appeal. The bulk surveillance issue is not one of those issues. The Minister ought to have sought leave from this Court in terms of rule 19. Lastly, the applicants submit that the Minister of State Security has not explained why it is in the interests of justice for a direct appeal to be heard.
 The applicants are correct to submit that the Minister of State Security ought to have sought leave to appeal. But in terms of section 173 of the Constitution, this Court is at liberty to determine its own process. A pragmatic approach is to entertain the appeal. The issue it raises was part of the same proceedings before the High Court. Determining it together with the confirmation proceedings will contribute towards a saving in time, costs and court resources; it simply makes practical sense. The appeal is here. All parties that wished to, have made submissions on it.
The issue to be determined is fairly simple, and no party will suffer any prejudice if we determine the merits at this stage. Requiring the Minister of State Security to go back and follow the normal appellate hierarchy or to bring a proper application for leave to appeal directly to us would be placing formalism ahead of what the interests of justice dictate. The appeal on the surveillance issue will thus be entertained. But this judgment must not be read to say that in comparable situations leave to appeal is not necessary. The determinant will be the circumstances of each case.
 Coming to the merits of the appeal, the Minister of State Security argues that the answer to the question at hand lies in a proper interpretation of section 2 of the National Strategic Intelligence Act (NSIA). She argues that in terms of this section, the state is empowered, subject to section 3 of the NSIA, to “gather, correlate, evaluate and analyse” various types of intelligence in order to identify any threat, potential or otherwise, to national security.
This provision, she argues, authorises the state to conduct bulk surveillance. Furthermore, RICA provides sufficient safeguards designed to ensure that this practice is not susceptible to abuse and unlawful invasion of privacy. The High Court disagreed with this interpretation, and held that the state’s practice of bulk surveillance is not authorised by the NSIA or any other law, and is thus unlawful.
 The impact of bulk surveillance is highlighted by the third amicus.
That amicus argues that:
- bulk surveillance entails the interception of virtually all internet traffic without a warrant or suspicion about the people whose communications are intercepted, and without statutory safeguards;
- no legal limits are placed on how data obtained through bulk surveillance is captured, copied, stored, analysed, or distributed;
- this unregulated, untargeted surveillance of all information is an extreme violation of the right to privacy;
- this violation is contrary to comparative and international law; and
- section 39(2) of the Constitution calls for an interpretation of the NSIA that avoids an extensive violation of the right to privacy and, therefore, does not permit unregulated bulk surveillance.
The effect of bulk surveillance, i.e. the interception of virtually all internet traffic without a warrant or suspicion about the people whose communications are intercepted, was not denied by the respondents.
 In a technologically developing society like ours, it is understandable that the state, through its intelligence services, may wish to use the latest technological means to ensure the safety of its citizens and uphold national security. These measures may involve the monitoring of communications – through these technological means – by citizens with others (whether citizens or not).
Without doubt, that monitoring constitutes the exercise of public power. And that power can be exercised only in a constitutionally compliant manner. In this regard, the principle of legality requires that any exercise of public power must have a basis in some law. Bulk surveillance must therefore have a legal basis. The question is: is section 2 of the NSIA on which the Minister of State Security relies that legal basis?
 Section 2 of the NSIA empowers certain state agencies, including the State Security Agency, on which the argument of the Minister of State Security focused, to “gather, correlate, evaluate and analyse” the following types of intelligence: domestic intelligence; crime intelligence; departmental intelligence; foreign intelligence; foreign military intelligence; and domestic military intelligence.
The section goes on broadly to describe various other functions, which include that the State Security Agency is authorised to:
- collect intelligence that may show threats to the Republic;
- supply such intelligence to other security agencies or governmental departments;
- liaise with other security agencies; and
- use the intelligence for analysis and other purposes.
 At best for the Minister of State Security, section 2 is ambiguous. Section 39(2) of the Constitution obliges us not only to give effect to the object of legislation, but also to seek a meaning of the relevant provisions of a statute that promotes the spirit, purport and objects of the Bill of Rights. Thus, we must interpret section 2 of the NSIA in a way that best promotes the right to privacy.
 With this in mind, it is significant that in RICA there is an express prohibition of communication interceptions without interception directions. To interpret section 2(1)(a) of the NSIA in the manner suggested by the Minister of State Security would be to undermine this prohibition.
Also, there is the principle of interpretation that says that where a provision is ambiguous, its meaning may be determined in light of other statutes on the same subject matter. This principle points away from the interpretation advocated by the Minister, particularly because it results in an intrusion into the privacy right that is expansive in its reach and most egregious in nature.
 Also, the section has shortcomings of another nature and these too impact negatively on the privacy right. This as well is an indication that the section ought not to be interpreted to authorise bulk surveillance. The section does not stipulate in clear, precise terms the manner, circumstances or duration of the collection, gathering, evaluation and analysis of domestic and foreign intelligence.
It merely broadly enumerates the “functions” of the State Security Agency, with no details as to the nuts and bolts of those functions. It also fails to set out the ambit of how these various types of intelligence must be captured, copied, stored, or distributed. It is thus not clear at all that section 2 provides for a practice that so significantly intrudes on the right to privacy.
 In sum, I am not convinced that the broad terms of section 2 serve as authorisation for the practice of bulk surveillance. The practice is thus unlawful and invalid, as there is no law that authorises it.
 Questions that arise from what was argued are whether – in the exercise of this Court’s remedial power under section 172(1)(b) – we should:
- limit the retrospective effect of the declaration of invalidity;
- suspend the declaration and, if so, for how long; or
- grant interim relief, in the event of a suspension of the declaration of invalidity, and
- if we do, what its nature should be.
 If the declaration of invalidity has full retrospective effect, that will render unlawful everything done under RICA pursuant to the grant of all previous interception directions. These should not be undone. That scrambled egg is incapable of being unscrambled. Indeed, this Court has held that “[a]s a general rule . . . an order of invalidity should have no effect on cases which have been finalised prior to the date of the order of invalidity”.
Suspension of declaration of invalidity
 There is no question that interceptions of communications in terms of RICA serve an important government purpose. That much is clear from the several purposes for which interception directions may be issued, which have been discussed above. The country’s surveillance system would be seriously dislocated if a declaration of invalidity were not suspended. To refocus our minds on what would be affected, let me render a collection of the purposes for which interception directions may be issued.
They may be issued if there are reasonable grounds to believe that:
- “a serious offence has been or is being or will probably be committed”;
- it is necessary to gather information concerning an actual threat to the public health or safety, national security or compelling national economic interests;
- it is necessary to gather information concerning a potential threat to the public health or safety or national security;
- the rendering of assistance to a foreign country in connection with or in the form of interception of communications is in accordance with an international mutual assistance agreement or is in the interests of South Africa’s international relations or obligations; or,
- lastly, it is necessary to gather information concerning property which is or could probably be an instrumentality of a serious offence or is or could probably be the proceeds of unlawful activities.
So, the dislocation of our surveillance system would have a grave impact on matters that are important to the country and its people. To avert this, we must suspend the declaration of invalidity.
 The High Court suspended its orders of invalidity for a period of two years to allow Parliament time to cure the various constitutional defects identified. The Minister of Justice requests that this period be increased to three years. He submits that a process to review RICA has already commenced although there is no Bill in place yet, and advises that the Executive requires time to proceed with its investigations and develop suitable remedial legislation.
To this end, he argues that three years is not inordinately long when the following factors, amongst others, are taken into account:
- the technical and complex nature of the legislation;
- the need for extensive consultation due to the contentious nature of the legislation and the obligations that it will impose on telecommunication service providers;
- the need for a careful consideration of developments and trends in other countries on the interception of communications; and
- the fact that the process of consultation and collation of comments received during the Bill drafting stage takes considerable time.
The Minister of Justice further submitted that three years will guard against the Executive having to return to this Court to request an extension.
 This is reasonable. Thus the declaration of invalidity must be suspended for three years.
 The High Court granted interim relief, the effect of which was to make substantial insertions into RICA, to be applicable during the period of suspension. The Ministers of Police and Justice argue that the doctrine of separation of powers militates against the interim relief because it strays into policy which is the preserve of the Executive.
And the Minister of Justice invokes Glenister II where Ngcobo CJ stated:
“Under our constitutional scheme it is the responsibility of the Executive to develop and implement policy. It is also the responsibility of the Executive to initiate legislation in order to implement policy. And it is the responsibility of Parliament to make laws. When making laws Parliament will exercise its judgment as to the appropriate policy to address the situation.”
 The respondents also emphasise that the interim relief will have unintended detrimental consequences on ongoing crime prevention measures by law enforcement agencies. They therefore submit that the expansive nature of the interim relief also falls within the terrain of legislating. That, according to them, is best left to the Legislature.
 The interim relief granted by the High Court and advocated by the applicants before us – entailing as it does, several insertions into RICA – is quite extensive. Those insertions do appear to be legislative in nature. That immediately raises the question whether it is within this Court’s province to grant relief of that nature. Courts are required – pursuant to declaring a legislative provision invalid – to balance the obligation to provide appropriate relief with the constitutional reality of the separation of powers principle.
Beyond that, the outer limits of a remedy are bounded only by considerations of justice and equity. That is indeed very wide. It may come in different shapes and forms dictated by the many and varied manifestations in respect of which the remedy may be called for. The odd instance may require a singularly creative remedy.
In that case, the court should be wary not to self-censor. Instead, it should do justice and afford an equitable remedy to those before it as it is empowered to.
In the words of Cameron J,
“the bogeyman of separation of powers concerns should not cause courts to shirk from [their] constitutional responsibility”.
 The infringement of the privacy right through the interception of individuals’ communications is egregiously invasive. The suspension of the declaration of invalidity means that this situation will persist for a while. And that is exacerbated by the fact that – because of the complexity and wide-ranging nature of the required revision – the period of suspension must be relatively long. So, the egregious invasion of the privacy right will persist for some time to come.
In those circumstances, justice and equity dictate that the effect of the intrusive violation of the privacy right be blunted by granting appropriate interim relief. Indeed, in Hoffman this Court held that in determining appropriate relief, it must
“carefully analyse the nature of the constitutional infringement, and strike effectively at its source”.
 Before considering the nature of interim relief, let me mention that I take the view that there must be no interim relief in respect of the independence issue. The objectionable facets in respect of this issue are not of such a nature that they may not be endured for the period of suspension.
 The High Court granted interim relief by reading-in a fairly substantial provision providing for post-surveillance notification. Essentially, the High Court made the giving of post-surveillance notification subject to exceptional circumstances. The Ministers of Police and Justice pointed out what they claimed to be unintended consequences of the provisions read-in by the High Court.
- their impact on other sections in RICA, which impact may necessitate a further reading-in;
- their application irrespective of whether an investigation is completed or still ongoing, or the subject is charged for the offences or not;
- uncertainty as to the manner in which the notification should take place; and
- the lack of clarity as to what “exceptional circumstances” should be considered for deferral of notification.
I do not quite see how – in a situation where a person is no longer under surveillance – post-surveillance notification may affect how RICA operates. This is a discrete addition that affects someone who – so to speak – is out of the surveillance system. How that system gets affected escapes me.
 The complaint about lack of clarity on exceptional circumstances does not have to arise. In this regard, there is guidance from foreign jurisprudence. In accordance with that jurisprudence, post-surveillance notification must be given as soon as that can be done without jeopardising the purpose of surveillance after surveillance has been terminated. I doubt that the relevant law enforcement officers will have any difficulties knowing whether notification after termination of surveillance will jeopardise the purpose of surveillance. After all they are best placed to know that.
 Of course, it can never be that the reason for the jeopardy will last forever with the result that, in a given case, a law enforcement officer can claim that at no stage will notification ever be given. In any event, what complaint the state may have in relation to the formulation of the basis for a temporary withholding of notification pales when regard is had to the egregious infringement at issue.
 It is thus appropriate to read-in provisions that provide for post-surveillance notification as the default position.
 I do not think it would be appropriate to order that – during the period of suspension – there must be automatic review.
- First, I have noted that automatic review is a safeguard worth considering; I did not conclude that it must be included in RICA.
- Second, the interim reading-in of an automatic review mechanism is complex.
For example, we do not know how many surveillance directions are executed over what period. Therefore, we have no idea what load of automatic reviews a designated Judge or designated Judges would have to attend to. Nor do we know what additional resources may be necessary. In the circumstances, I do not think it appropriate to allow automatic review in the interim.
Lawyers and journalists
 Save for some minor adjustments, the interim relief granted by the High Court in respect of interception or surveillance directions affecting practising lawyers and journalists is appropriate.
The rest of the issues
 In respect of the remaining issues, I take the view that interim relief should not be granted.
 The High Court held that “[i]n keeping with the character of the controversy and the conventions in this genre of litigation, there shall be no order as to costs”. The applicants seek their costs and are appealing against this. They claim that this is a misdirection because if applicants are successful in a constitutional challenge against the state, they should be awarded their costs.
 The principles on costs in constitutional litigation were established in Biowatch.
This Court held that—
“the general rule for an award of costs in constitutional litigation between a private party and the state is that if the private party is successful, it should have its costs paid by the state, and if unsuccessful, each party should pay its costs.”
 This Court is generally reluctant to interfere in costs orders of the courts below—
“unless it is satisfied that the discretion was not exercised judicially, the discretion was influenced by wrong principles, or a misdirection on the facts, or the decision reached could not reasonably have been made by a court properly directing itself to all the relevant facts and principles. There must have been a material misdirection on the part of the lower court”.
 These are such exceptional circumstances. It is unclear which genre of litigation the High Court is referring to, but – on the authority of Biowatch – its decision on costs is based on a material misdirection. According to Biowatch the applicants ought to have been awarded their costs against all the respondents that opposed the application before the High Court. In this Court as well the applicants are entitled to their costs. Only the Ministers of State Security and Police can be said to have really opposed the confirmation application. They must pay the costs in this Court.
 The following order is made:
1. The appeal by the Minister of State Security is dismissed with costs, including the costs of two counsel.
2. The appeal by the Minister of Police is dismissed with costs, including the costs of two counsel.
3. The appeal by the applicants against the costs order granted by the High Court, Gauteng Division, Pretoria (High Court) is upheld with costs, including the costs of two counsel.
4. The High Court’s order referred to in paragraph 3 is set aside.
5. The Minister of Justice and Correctional Services, the Minister of State Security, the Minister of Defence and Military Veterans, the Minister of Police, the Office for Interception Centres, the National Communications Centre and the State Security Agency must pay the applicants’ costs of the application before the High Court, such costs to include the costs of two counsel.
6. The declaration of unconstitutionality by the High Court is confirmed only to the extent that the Regulation of Interception of Communications and Provision of CommunicationRelated Information Act 70 of 2002 (RICA) fails to—
(a) provide for safeguards to ensure that a Judge designated in terms of section 1 is sufficiently independent;
(b) provide for notifying the subject of surveillance of the fact of her or his surveillance as soon as notification can be given without jeopardising the purpose of surveillance after surveillance has been terminated;
(c) adequately provide safeguards to address the fact that interception directions are sought and obtained ex parte;
(d) adequately prescribe procedures to ensure that data obtained pursuant to the interception of communications is managed lawfully and not used or interfered with unlawfully, including prescribing procedures to be followed for examining, copying, sharing, sorting through, using, storing or destroying the data; and
(e) provide adequate safeguards where the subject of surveillance is a practising lawyer or journalist.
7. The declaration of unconstitutionality in paragraph 6 takes effect from the date of this judgment and is suspended for 36 months to afford Parliament an opportunity to cure the defect causing the invalidity.
8. During the period of suspension referred to in paragraph 7, RICA shall be deemed to include the following additional sections:
“Section 23A Disclosure that the person in respect of whom a direction, extension of a direction or entry warrant is sought is a journalist or practising lawyer
(1) Where the person in respect of whom a direction, extension of a direction or entry warrant is sought in terms of sections 16, 17, 18, 20, 21, 22 or 23, whichever is applicable, is a journalist or practising lawyer, the application must disclose to the designated Judge the fact that the intended subject of the direction, extension of a direction or entry warrant is a journalist or practising lawyer.
(2) The designated Judge must grant the direction, extension of a direction or entry warrant referred to in subsection (1) only if satisfied that it is necessary to do so, notwithstanding the fact that the subject is a journalist or practising lawyer.
(3) If the designated Judge issues the direction, extension of a direction or entry warrant, she or he may do so subject to such conditions as may be necessary, in the case of a journalist, to protect the confidentiality of her or his sources, or, in the case of a practising lawyer, to protect the legal professional privilege enjoyed by her or his clients.”
“Section 25A Post-surveillance notification
(1) Within 90 days of the date of expiry of a direction or extension thereof issued in terms of sections 16, 17, 18, 20, 21 or 23, whichever is applicable, the applicant that obtained the direction or, if not available, any other law enforcement officer within the law enforcement agency concerned must notify in writing the person who was the subject of the direction and, within 15 days of doing so, certify in writing to the designated Judge, Judge of a High Court, Regional Court Magistrate or Magistrate that the person has been so notified.
(2) If the notification referred to in subsection (1) cannot be given without jeopardising the purpose of the surveillance, the designated Judge, Judge of a High Court, Regional Court Magistrate or Magistrate may, upon application by a law enforcement officer, direct that the giving of notification in that subsection be withheld for a period which shall not exceed 90 days at a time or two years in aggregate.
9. The Minister of Police and the Minister of State Security must pay the applicants’ costs in this Court, including the costs of two counsel.
The following explanatory note is provided to assist the media in reporting this case and is not binding on the Constitutional Court or any member of the Court.
On 4 February 2021 at 10h00 the Constitutional Court handed down judgment in an application to confirm the declaration by the High Court of South Africa, Gauteng Division, Pretoria (High Court) that the Regulation of Interception of Communications and Provision of Communication-Related Information Act 70 of 2002 (RICA) is unconstitutional, to the extent that it fails to provide adequate safeguards to protect the right to privacy, as buttressed by the rights of access to courts, freedom of expression and the media, and legal privilege.
The applicants, AmaBhungane Centre for Investigative Journalism NPC and Mr Stephen Sole – a journalist who had been the subject of state surveillance – approached the High Court on the basis of a number of constitutional challenges to RICA. The High Court upheld the following challenges:
(i) RICA makes no provision for a subject of surveillance ever to be notified that she or he has been subjected to surveillance (notification issue);
(ii) RICA permits a member of the Executive unfettered discretion to appoint and renew the term of the designated Judge (the functionary responsible for issuing directions for the interception of private commnications), and thus fails to ensure the independence of the designated Judge (independence issue);
(iii) RICA lacks any form of adversarial process or other mechanism to ensure that the intended subject of surveillance is protected in the ex parte application process (ex parte issue);
(iv) RICA lacks adequate safeguards for examining, copying, sharing, sorting through, using, destroying and/or storing the surveillance data (management of information issue); and
(v) RICA fails to provide any special circumstances where the subject of surveillance is a journalist or practising lawyer (practising lawyers and journalists issue). RICA was accordingly declared unconstitutional to the extent of these failures.
The declaration of invalidity was suspended for two years to allow Parliament to cure the defects. Interim relief, in the form of reading-in, was granted in respect of the notification issue, the independence issue and the practising lawyers and journalists issue.
The applicants supported the bulk of the reasoning and findings of the High Court, save for that relating to costs, which they appealed. The Minister of Police partially appealed the judgment and orders of the High Court; and the Minister of State Security appealed the whole judgment and orders of the High Court. The Minister of Justice did not appeal the High Court’s declaration of invalidity or oppose the application for confirmation, but made submissions to assist the Court. The rest of the respondents did not participate in the Constitutional Court.
In a majority judgment penned by Madlanga J (Khampepe J, Majiedt J, Mathopo AJ, Mhlantla J, Theron J, Tshiqi J and Victor AJ concurring), the Constitutional Court held that interception and surveillance of an individual’s communications under RICA is a highly invasive violation of privacy, and thus infringes section 14 of the Constitution.
The Court accordingly considered whether this limitation is reasonable and justifiable under section 36(1) of the Constitution.
The Court acknowledged the constitutional importance of the right to privacy, which is tied to dignity. It further accepted the importance of the purpose of state surveillance, which is to investigate and combat serious crime, guarantee national security, maintain public order and thereby ensure the safety of the Republic and its people.
However, in light of the egregiously intrusive nature of the limitation, the Court held that the question to answer is: is RICA doing enough to reduce the risk of unnecessary intrusions? In other words, are there safeguards that acceptably minimise the trampling of the privacy right, and thus meet the reasonableness and justifiability standard?
On the notification issue, the Minister of Police, who appealed the High Court’s order in this regard, argued for the continued prohibition of all post-surveillance notification. The Constitutional Court held that such a blanket prohibition facilitates the abuse of interception directions, which are applied for, granted and implemented in complete secrecy.
Even if a direction ought not to have been granted, the subject will never know and is thus denied the opportunity to seek legal redress for the violation of her or his right to privacy. This renders the rights guaranteed by sections 34 and 38 of the Constitution to approach a court to seek appropriate relief for the infringement of the right to privacy illusory, and promotes impunity. Post-surveillance notification would serve a purpose comparable to less restrictive means. The Court concluded that post-surveillance notification should be the default position. RICA was held to be unconstitutional to the extent that it fails to provide for notifying the subject of surveillance of her or his surveillance as soon as notification can be given without jeopardising the purpose of surveillance after it has been terminated.
Before adjudicating the independence issue, the Court considered whether RICA empowers the Minister of Justice to appoint the designated Judge at all. RICA has as its centrepiece this designated Judge, who may authorise surveillance of communications in limited circumstances.
However, aside from the definition of “designated Judge”, which refers to a judge “designated by the Minister of Justice for the purposes of [RICA]”, there is no provision expressly empowering the Minister to appoint the designated Judge. This constitutional issue was not pleaded, but raised by the Court of its own accord.
The Court drew a distinction between a primary implied power and an ancillary implied power. While the latter is necessary for, and implied by, an unquestionably existing power to be exercised, the former is a power to do something required to be done in terms of an Act, and which does not owe its existence to some other power. Reverting to RICA, the Court held that, reading the definition and the provisions on the functions of a designated Judge together, and considering that the main purpose of RICA cannot be achieved without a properly designated Judge, the Minister’s power to designate a Judge was implied in the definition of “designated Judge”.
The Court then turned to the independence issue, which was founded on the grounds that: RICA failed to prescribe or limit the designated Judge’s term of office, making it possible for the Minister to make indefinite reappointments; each term was for a duration determined at the whim of the Minister; and appointments of designated Judges were exclusively made by a member of the executive in a non-transparent manner, in that there was no role for the Judicial Service Commission, Parliament or the Chief Justice. The Court held that the open-ended discretion in respect of appointments and their renewal could raise a reasonable apprehension that the independence of the designated Judge may be undermined by external interference by the Executive. As a result, RICA does not allow the designated Judge an adequate level of structural, operational or perceived autonomy. The Court declared RICA unconstitutional to the extent that it fails to ensure adequate safeguards for an independent judicial authorisation of interception.
On the ex parte issue, the Court held that means to temper the effects of the clandestine, one-sided nature of the process do exist, and constitute less restrictive means to achieve the purpose of surveillance. RICA was thus found to be unconstitutional to the extent that it lacks sufficient safeguards to address the fact that interception directions are sought and obtained ex parte. The Court left the choice of what measures are most suitable to Parliament.
The Court then considered the management of information issue. The applicants’ concern was that the lack of regulation of how intercepted information is handled, stored and eventually destroyed exposes subjects of interceptions to even more aggravated intrusions into their privacy. The Court considered various provisions that the Minister of State Security argued fulfil this role. The Court held that these provisions do not prescribe the relevant procedures, and that they allow the Director of the Office for Interception Centres an unacceptably unbounded discretion to regulate the management of information. This results in unnecessarily egregious intrusions into the privacy of the subjects of interceptions. The Court thus declared RICA unconstitutional to the extent that it fails adequately to prescribe procedures to ensure that data obtained pursuant to the interception of communications is managed lawfully and not used or interfered with unlawfully.
In dealing with the lawyers and journalists’ issue, the Court acknowledged that the confidentiality of journalists’ sources is protected by the rights to freedom of expression and the media. In relation to the confidentiality of lawyers’ communications, the Court accepted that legal professional privilege is an essential part of the rights to a fair trial and fair hearing. These rights were found to weigh in favour of special consideration being given to the importance of the confidentiality of lawyer-client communications and journalists’ sources, in order to minimise the risk of infringement of this confidentiality. RICA’s failure to do so rendered it unconstitutional.
The Court then dealt with the Minister of State Security’s appeal regarding the legality of the bulk communications surveillance. The High Court had declared bulk surveillance unlawful. The Constitutional Court dismissed the appeal.
It held that section 2 of the National Strategic Intelligence Act 39 of 1994 is ambiguous, and should thus be interpreted in a manner that best promotes the right to privacy, and does not contradict RICA’s prohibition of communication interceptions without interception directions. The broad terms of section 2 thus do not authorise the practice of bulk surveillance. The practice was declared unlawful and invalid.
Having declared RICA unconstitutional, the Court limited the retrospectivity of its declaration of invalidity. It further suspended its declaration of invalidity for three years, as requested by the Minister of Justice, to allow Parliament adequate time to proceed with its investigations and develop suitable remedial legislation. Since the infringement of the privacy right is egregiously intrusive, and the period of suspension is relatively long, the Court deemed it necessary to grant interim relief in respect of the notification issue, and the lawyers and journalists issue.
Finally, the Court considered the applicants’ appeal on costs. Despite their success, the High Court had denied the applicants their costs. It decided that there should be no order as to costs. The Constitutional Court acknowledged that costs awards of courts below should be interfered with only in exceptional circumstances. However, since the High Court had failed to apply established principles on costs in constitutional litigation, these were exceptional circumstances. The applicants – private parties – had been successful in constitutional litigation against the state. They were thus entitled to their costs both in the High Court and Constitutional Court.
The second judgment penned by Jafta J (Mogoeng CJ concurring) agreed with the first judgment save for the finding in the first judgment that RICA empowers the Minister to designate a judge for the purposes of determining applications for authorisation to intercept private communications and also to perform other functions. The second judgment held that the definition in section 1 of RICA does not include a provision that the Minister has the power to designate but merely defines the meaning of the term “designated judge”.
In addition to the reasons set out in the first judgment regarding the unconstitutionality of Chapter 3 of RICA, the second judgment held that the limitation to the right to privacy is unjustifiable and unreasonable, absent a lawful and valid designation of a Judge.
Consequent upon the finding that RICA does not empower the Minister to designate a judge, the second judgment held that the suspension of the declaration of invalidity proposed as a remedy in the first judgment is inappropriate as it will not cure the problem of the lack of power to designate. This kind of problem can only be remedied by Parliament granting the Minister the relevant power. As a result, the second judgment concluded that the appropriate remedy should have been to simply declare the impugned provisions invalid with no additional remedies.