Skip to content

The Intelligence Gap: what any review into alleged failings after recent terrorist attacks needs to address and who needs to address it.

Three terrorist attacks in England in as many months, the most recent on 3 June starting on London Bridge and ending in Borough market. In each case, there has been a suggestion that some of those involved were known to the intelligence services. Inevitably, there have been calls for a review. This is not the first time it has happened. The 7/7 ring-leader, Mohammad Sidique Khan was known to MI5 before the bombings and although the deputy Coroner cleared the security service of any wrong-doing for not paying him more attention, she recommended it improved the way it recorded decisions relating to target-assessment. This was a particularly significant recommendation since it was made in the context of what the Security Service had done since the bombings. Similarly, Lee Rigby’s murderers had been known to MI5 and one of them, Michael Adebolajo, had alleged that MI5 had persistently questioned him and had even attempted to recruit him. An inquiry by the Intelligence and Security Committee (ISC) found that the murder would not in any event have been avoided. So is it a case, as Joseph Conrad wrote in A Secret Agent, his prescient novel about a plan to bomb Greenwich Observatory, that “there are more kinds of fools than one can guard against” or as Paddy Ashdown tweeted, “more a lack of resources than dark spaces”?

What is clear is that the intelligence services have greater powers on the statute book than ever before. This includes the product of what became known as mass surveillance but is referred to in the legislation inelegantly as bulk powers. Bulk powers extend over the range of telecommunications and postal surveillance powers available to MI5, MI6 and GCHQ as well as a range of other public authorities. This includes the interception of communications, acquisition and retention of communications data, hacking and exploitation of what are known as bulk personal datasets. The legal basis for their use is now set out in the vast Investigatory Powers Act 2016 or the IPA. During its evolution, questions were asked about the “operational case” for their use. The then Independent Reviewer of Terrorism Legislation, David Anderson QC, was tasked to carry out a review and report. He concluded that there was a proven operational case in respect of their use with the exception of bulk hacking (which had yet to be made out) but added:

This Report has declared the powers under review to have a clear operational purpose.  But like an old-fashioned snapshot, it will fade in time.  The world is changing with great speed, and new questions will arise about the exercise, utility and intrusiveness of these strong capabilities.  If adopted, my recommendation will enable those questions to be answered by a strong oversight body on a properly informed basis.

David Anderson, tweeted earlier this week to say, by reference to his report on bulk powers, that it’s hard to argue their use is ineffective. This led to Professor Martin Scheinin, the former UN Special Rapporteur on human rights and counter-terrorism to say that, “if human intelligence gives you real targets, then some of the cases cited [by Anderson] might work and be justified as targeted surveillance”. This was a striking exchange: in less than two tweets, respected experts had encapsulated the modern surveillance dichotomy. The metaphor that has been used was wrong – it was not about finding a needle in an ever-larger haystack – its size was irrelevant if someone could point to it and say, “over here is where the needle can be found”. An effective counter-terrorism strategy, as the experience of Northern Ireland bears testimony, embodies both human intelligence – in all its guises (from acquisition and revelation of intelligence, its exploitation to infiltration) – and technical surveillance, including bulk powers, properly defined and above all, rigorously and independently scrutinised. The pressing issue for the intelligence services is not the nature of its powers but whether how their use of them has changed as they have increased and whether this has created an intelligence gap.

Any review needs to resolve this dichotomy, not engage in the futility of where to place blame. But it is not just the nature of the review that needs to be reframed but also who should have responsibility for it. Under sections 235 and 236 of the IPA, the new Investigatory Powers Commissioner may be asked by either the Prime Minister or the ISC to make a report into any matter. Later this week, once the country has a new government, it would be open to whoever leads it to direct an inquiry: its terms of reference defined, perhaps for the first time but no worse for it, by the tweets of two leading experts in their fields. It will also be a test of whether it the Commissioner will provide the strong oversight called for by David Anderson. If there is a failure to address this most pressing of issues there is a risk that more misguided fools may walk, as Conrad forewarned, “unsuspected and deadly, like a pest in the street full of men”.  

SPIES, LIES AND FIVE EYES: why GCHQ could not and would not have carried out surveillance on the presidential candidate.

In order for GCHQ to carry out technical surveillance on Donald Trump it would have needed a warrant signed personally by the Secretary of State. Obtaining a warrant is governed by section 5 of the Intelligence Services Act 1994 (ISA). This provides, insofar as relevant, that in order to obtain a warrant the Foreign Secretary must personally think it is necessary for the action to be taken by the intelligence services, including GCHQ. This must relate to the carrying out of any function which falls within section 3(1)(a) of the ISA (and includes what is commonly referred to as “wire tapping”). It could theoretically include the interception of communications. In those circumstances sections 5 and 8(4) of the Regulation of Investigatory Powers Act 2000 (RIPA) would apply.

 A warrant – under either ISA or RIPA – could only be signed in the present context if it was in the interests of national security, having regard to the UK’s defence and foreign policies or in the interests of the economic well-being of the United Kingdom in relation to the actions or intentions of persons outside the British Islands. Is it possible that the activities of a lawfully nominated presidential candidate could reach this high threshold? This is so extremely unlikely that it is fanciful but even if it could be conceived, there are other difficulties.

The principal difficulty is what has become known as the Five Eyes Agreement. It has been suggested in some press reports that this is in fact the gateway used by the former President to facilitate surveillance of Trump. “Judicial analyst”, academic Andrew Napolitano and commentator for, amongst others, Fox News reported that:

“Three intelligence sources have informed Fox News that President Obama went outside the chain of command. He didn’t use the NSA. He didn’t use the CIA. He didn’t use the FBI, and he didn’t use Department of Justice. He used GCHQ. What the heck is GCHQ? That’s the initials for the British spying agency. They have 24/7 access to the NSA database. So by simply having two people go to them saying, ‘President Obama needs transcripts of conversations involving candidate Trump, conversations involving president-elect Trump,’ he’s able to get it, and there’s no American fingerprints on this”

There is an obvious flaw in this analysis, which is that it suggests GCHQ are accessing existing intelligence via the NSA database (i.e. material the US intelligence agencies already have on Trump). Such material would have “American fingerprints” all over it and would in any event need to be authorized. Under the Five Eyes agreement, President Obama wouldn’t have direct access to GCHQ – this is via secure liaison – but in any event the direction of traffic is going the wrong way. For this hypothesis to have any credence it would have to be the UK who identified Trump as a threat, obtained a warrant and then sought under the Five Eyes agreement express approval from the US authorities to get on with their surveillance activities. Quite apart from the absence of any legal basis existing (since the grounds could not objectively be made out) it would be diplomatically dynamite. Legal advice to the British Foreign Secretary – this would be routinely sought – would never have sanctioned such a course. The Prime Minister would have been briefed. Moreover, Napolitano’s theory that the former President managed to “leap-frog” his own intelligence apparatus is absurd. For his premise to work, US intelligence, far from not knowing anything about it, would have had to sanction it. If there were any truth in it, we would be facing a diplomatic incident of unparalleled seriousness. That does not appear to be the case. The greater risk to diplomacy is an attempt to prop up one fiction – that US spy agencies “tapped” Trump – by the invention of an even greater one, that British spy agencies did so.

Reflections on Love v National Crime Agency

Officers from the National Crime Agency (NCA) arrested Lauri Love during a search of his parent’s home on 25 October 2013. The alleged offences were under s 1(1) and (3) of the Computer Misuse Act 1990 (CMA 1990). He was released on police bail but in or around July 2014 he was shown a letter by his MP from the Crown Prosecution Service stating that no further action was being taken against him.

In February 2014 the NCA served a notice under s 49 of the Regulation of Investigatory Powers Act 2000 (RIPA 2000). In summary such a notice required disclosure by Mr Love to the NCA of the passwords relating to the devices they had seized. Evidently Mr Love failed to comply with the notice. However the NCA did not pursue the matter further.

The application

Mr Love brought an application under s 1 of the Police Property Act 1897 for the return of computer and storage devices seized during the search. During the course of these proceedings, the NCA sought an order for disclosure of the passwords. This request quite properly concerned a number of high profile commentators, including David Allen Green and Ryan Gallagher.

In the event, the District Judge ruled that to order disclosure would circumvent the provisions of RIPA 2000. However the case gives to other issues that this blog explores (note, I am not being declarative about the points made).

The NCA was created by s 1 of the Crime and Courts Act 2013 to investigate serious and organised crime. Offences under the MCA 1990 certainly have the potential to fall within what most would consider serious and organised crime. The NCA’s permitted purposes are broad and include the investigation of any crime (whether serious and organised or not) but clearly questions could properly be asked about the allocation of resources if it was routinely investigating offences that were less than serious and organised. By the time the Police Property Act application was considered it appears the investigation into Mr Love had changed somewhat, the NCA identifying only that it suspected he had names and passwords of users of the online magazine Police Oracle and films allegedly copied in breach of copyright. These are significantly less serious than offences under MCA 1990 and are often dealt with as civil, not criminal, matters.

Section 49 RIPA 2000

Under s 49 RIPA 2000, the NCA would ordinarily need to get appropriate permission from a circuit judge unless it was covered by one of the exemptions (it is not clear whether an exemption applied in Mr Love’s case – but the s 49 notice followed the arrest by 5 months – so this seems unlikely). It is important to remember that a judge is not ordering disclosure of the passwords but rather giving permission for the service of the notice. If the recipient of the notice fails to comply he or she can be charged with an offence under s 53 of RIPA 2000, which carries up to five years imprisonment on conviction on indictment.

Two serious issues arise then from the attempt by NCA to secure the passwords in the course of the Police Property Act proceedings. The first is that they were seeking an order that went beyond the powers of a judge under RIPA 2000: s 49 would not require disclosure of the passwords, merely subject the recipient to the risk of a custodial sentence if found guilty of failing to comply. The second is that they were asking for the order from a District Judge not a Circuit Judge and is not clear the criteria was met for them to do so.

There are no particular rules as to the form of an application for appropriate permission to serve a s 49 Notice under RIPA. It is a question of substance and the NCA’s application appears to have at least included, for all intents and purposes, an application for appropriate permission under s 49. Since they already had secured a notice but had chosen not to bring proceedings alleging an offence had been committed the application was on the face of it duplicitous (in the legal sense) and abusive.

The Code of Practice

The Code of Practice on Investigation of Protected Electronic Information (encryption) requires at paragraphs 3.36-4.42 that the NCA must be satisfied it is necessary to seek a notice and has to consider, by way of one example, “the benefit to the investigation…” (the link to the Code is at the time of writing unavailable). It is not clear there remained an “investigation” (since it has been referred to the CPS and a decision has been taken not to charge) so it is questionable that NCA could be so satisfied. It could be argued the file remains open but insofar as offences are capable of now being identified, these appear to relate to copyright infringement. Moreover there is no evidence that the test of necessity and proportionality were considered. The NCA have previously fallen foul of the application of this test in the recent case of Chatwani (discussed here).

Section 65, RIPA 2000

There is another important point that arises and it is the question of the jurisdiction of the criminal courts to hear the application. This is governed by s 65 RIPA 2000. Mr Love was running a HRA 1998 argument (that any order would breach s 6 of 1998 Act as it would be a violation of Article 8 of the European Convention on Human Rights). Under s 65(2)(a) the Investigatory Powers Tribunal is the “only appropriate tribunal for the purposes of s 7, HRA” (incompatibility: see for an example of the operation of these provisions the Supreme Court case of A v B). The proceedings have to fall within s 65(3). These include “proceedings relating to the taking place in challengeable circumstances of any conduct falling within subsection (5)”. Subsection (5) includes at (e) “the giving of a notice under s 49…”. It is not clear but the jurisdiction of the Magistrates’ Court may have been ousted as a result.

© Simon McKay (2016). Please ask for my permission before using or copying the material contained in this blog. If you do use it, attribute it to me with a link to the original content where possible. This is opinion, not legal advice.


GCHQ and its Chamber of Secrets

It was reported on Sunday that GCHQ intervened to prevent the leak of the sixth instalment of JK Rowling’s Harry Potter series. In response the spy agency said that it “didn’t comment on [its] defence against the dark arts”. The story was no doubt intended to demonstrate that United Kingdom’s third intelligence service has a sense of humour but to many it may have been deeply insensitive.

Disclosures made by any of the intelligence services are tightly regulated by statute. The Security Service Act 1989 governs disclosures by the Security Service (MI5) and the Secret Intelligence Service (MI6) and GCHQ, by the Intelligence Services Act 1994. The provisions in relation to the disclosures each makes are in identical terms but those in respect of GCHQ are as follows:

4 The Director of GCHQ.

(1) …

(2) The Director shall be responsible for the efficiency of GCHQ and it shall be his duty to ensure—

(a) that there are arrangements for securing that no information is obtained by GCHQ except so far as necessary for the proper discharge of its functions and that no information is disclosed by it except so far as necessary for that purpose or for the purpose of any criminal proceedings

The government places great emphasis on the strict application of these – what it calls – “the statutory provisions on disclosure of information”. Indeed in a witness statement provided to the Pitchford Inquiry on Undercover Policing, a cabinet office witness has correlated the provisions with the need to avoid a “hostile individual or group” thwarting an investigation or operation or to discover the methods, capabilities or techniques used by our intelligence services. This is in addition to compromising the safety of agents or prospective agents whose lives could be “put at personal risk” if the provisions in relation to disclosure were less than strictly applied. For these reasons, says the Cabinet Office, the same principles are “also likely to apply to undercover operations conducted by the police”. Indeed, this is the entire premise underpinning the Metropolitan Police’s argument to rely on the policy of Neither Confirm Nor Deny (“NCND”, explained here) in the Undercover Policing Inquiry, which is investigating, amongst other things, the practice by some officers of sleeping with activists or their associates.

Just to get a flavour of how serious the Metropolitan Police (eventually) considered this practice it may be worth recalling the apology the victims received from assistant commissioner Hewitt after attempts to strike out the legal action failed:

“I acknowledge that these relationships were a violation of the women’s human rights, an abuse of police power and caused significant trauma. I unreservedly apologise on behalf of the Metropolitan police service. I am aware that money alone cannot compensate the loss of time, their hurt or the feelings of abuse caused by these relationships.

“Most importantly, relationships like these should never have happened. They were wrong and were a gross violation of personal dignity and integrity”.

The principal basis of the assertion to NCND is therefore interwoven in the legislation that strictly regulates the disclosures that can properly be made by the intelligence services. It is being used to seek an order that much of the public inquiry be heard in secret a departure from the established principle of open justice.

The provisions of s 4(2) (a) also appear in the report of Sir Peter Gibson in connection with his report into the intelligence sharing that took place prior to the Omagh bombing. He said this:

“Once intercept material reached RUC HQ and Special Branch South, any further publication and release of that material, even to another part, or other members, of Special Branch, was subject to strict conditions imposed by GCHQ designed to achieve a balance between providing support to customers like Special Branch and protecting GCHQ’s capabilities, sources and methods. GCHQ also sought to ensure compliance with its legal obligations, in particular that required of the Director of GCHQ by section 4(2)(a) of the Intelligence Services Act 1994, viz to ensure that no information was disclosed by GCHQ except so far as necessary for the proper discharge of its functions or for the purpose of any criminal proceedings. If those persons within the RUC HQ and Special Branch South who received intelligence from GCHQ wanted to disseminate it within the RUC or even within Special Branch a set procedure had to be followed. GCHQ’s permission had to be sought for the use of intelligence in a ‘sanitised’ form, that is without revealing its source, to carry out some authorised action. Thus the release of intelligence intended to be acted on by other officers had to be requested of GCHQ and a form of words cleared with GCHQ”.

The reaction to organs of the state asserting secrecy is one of cautious if not reluctant acceptance that “it” knows best and in serving the greater good of our overall security some tolerance of secrecy is necessary. However this timid balance depends on public confidence that such restrictions placed on the public’s right to know is exercised appropriately and only when strictly necessary. Revelations that spy agencies misuse it, even when it relates to a fictional, though highly popular, orphaned wizard, is insensitive and undermines the great tragedy and intrusive violations that some have had to endure in the absence of full accountability. Unless the state is prepared to open a Pandora’s box, it should keep its chamber of secrets firmly closed.

Neither Confirm Nor Deny and Restriction Orders in Public Inquiries

[The subject of this blog forms the basis of a paper due to be published later this year. As a result it focuses on a discrete issue touching on the policy of Neither Confirm Nor Deny (NCND) in the context of public inquiries.]

The Pitchford Inquiry into Undercover Policing is examining the legal principles governing the application of NCND in the context of s 19 of the Inquiries Act 2005 (“the 2005 Act”; it being an inquiry ordered under the provisions of that statute). The state parties have filed submissions and a witness statement has been filed by the Cabinet Office to assist the inquiry.

All of the submissions deal with the key cases, discussed in an earlier blog here.

In public inquiries, restriction notices (made upon a Minister giving to the chairman a notice) or orders (determined by the chairman) can be made under s 19 of the 2005 Act. As the statement from the Cabinet Office makes clear “police forces…are not part of any government department or agency”, so notices are unlikely to arise during the course of the inquiry (although the Metropolitan Police has intimated it will seek an order as opposed to a notice but it is not immediately clear on what basis it could seek a notice from a Minister).

Restriction orders can only be made where the chairman of the inquiry considers it conducive to the fulfilling of its terms of reference or to be necessary in the public interest having regard to four matters: (a) the effect on inhibiting public concern; (b) avoiding or reducing any harm or damage that might be caused without an order being made; (c) respecting confidentiality; and (d) where not making the order may cause delay or impair the efficiency or effectiveness of the inquiry or result in additional costs. The matters are not exhaustive, so it is open to the inquiry to also consider the significance of NCND if it applies on the facts of the case.

The approach that ought to be taken is not easily identifiable. This blog sets out a possible route:

  1. Does the application of NCND arise in principle on the facts of the individual case?

1.1     If no, no order should be made, subject to (4) below.

1.2     If yes, consider (2)

  1. Should NCND apply having regard to the facts of the case? If not applied will it:

2.1     Thwart any on-going investigation or operation?

2.2     Reveal methods, capabilities or techniques, not already in the public domain and avowed, that if disclosed, would damage the public interest?

2.3    Jeopardise the willingness of future police officers to volunteer for undercover work?

2.4     Result in a real and immediate threat to life or injury?

2.5     If no, no order should be made, subject to (4) below.

2.6     If yes, consider (3).

  1. Are there any reasons why NCND should not apply?

3.1     Has there been any official confirmation that would negate the operation of the policy?

3.2     Is there evidence that the officer engaged in conduct that would amount to “crime or inequity”: Gartside v Outram (1857) 26 LJ Ch (NS) 113 such as to negate the confidential nature of his role?

3.3     Is there evidence that the officer is seeking to rely on the policy in order to frustrate the objectives of the inquiry?

3.4     Is there a risk justice might miscarry if the policy was applied?

3.5     Is it obvious to the reasonable observer that the officer was deployed in an undercover capacity?

3.6     Is there some other reason that needs to be weighed in the balance in determining whether the policy should be applied?

  1. The application of s 19 of the 2005 Act (assuming no issues of delay, efficiency or effectiveness or cost arises)

4.1     Is there a real and immediate risk of death or injury to the witness if no order is made and if so, will any order avoid or reduce this risk (bearing in mind this is a high threshold and that any threat must be supported by objective verification: Re Officer L [2007] UKHL 36)?

4.2     Has the witness been given any assurances about the confidentiality of the information they have provided or may provide to the inquiry and if so will any order respect any conditions as to confidentiality?

4.3     Will the making of the order inhibit the allaying of public concern?

4.4     On balance is an order necessary in the public interest having regard to the matters set out at 4.1 to 4.3 above?

4.5     If no, does the application of the policy of NCND, if made out, justify the making of an order?

4.6     If no, no order should be made

4.6     If yes, an order should be made but consideration should then be given to how its terms can interfere with the principle of open justice to the least extent possible.

This is a suggested route, applying the identifiable principles. It is not reaching any view on what the terms of any orders should be.

© Simon McKay (2016). Please ask for my permission before using or copying the material contained in this blog. If you do, attribute it to me with a link to the original content where possible. This is opinion, not legal advice.

In a word: privacy and the Investigatory Powers Bill

The Intelligence and Security Committee (ISC) in its report in response to the draft Investigatory Powers Bill (the Bill) said, “Overall, the privacy protections are inconsistent and in our view need strengthening. We recommend that an additional Part be included in the new legislation to provide universal privacy protections, not just those that apply to sensitive professions”. In less than a month, the revised Bill was published. In response to the ISC’s recommendation, the parliamentary draftsperson added under the heading “Part 1” the heading “General Privacy Protections” but failed to change the content of Part 1 in any other material way. The failure to do so caused astonishment. David Allen Green at Jack of Kent exposed the cynicism in a blog called “Privacy is Surveillance”, Buzz Feed’s James Ball tweeted dryly “ISC said ‪#IPBill needed a new section to “provide universal privacy protections”. Instead, they changed a heading” and NSA whistleblower Edward Snowden, who exposed the spectre of unlawful mass surveillance by state agencies on a global scale and catalyzed the current reform tweeted sardonically “Headline: “Revised bill adds privacy safeguards.” Reality: Govt changed one header”.

The need for universal privacy protection in the context of covert policing resources is made clear when a typical surveillance scenario is considered: surveillance of a telephone conversation.

If the conversation is intercepted, it needs a Ministerial certificate. Interception has consistently been regarded as “highly intrusive activity”: see the government’s Consultation on Interceptions in 1999. If it takes place without lawful authority it is a criminal offence: s 1, Regulation of Investigatory Powers Act 2000 (RIPA). The content of the conversation is inadmissible and its disclosure is a criminal offence: ss 18 and 19 RIPA.

However, if the conversation takes place on a “hands-free” telephone within a car, the authorisation regime is different. A combined interference with property and intrusive authority is required. This is self-authorised at Chief Constable level but requires approval from a Surveillance Commissioner. The content of the conversation is admissible: R v Alsopp [2009] EWCA Crim 462.

If a person present listening to the conversation (or indeed recording it) is a Covert Human Intelligence Source (and informer or undercover officer) the authority required is internal – superintendent or above – and not subject to the approval regime. The Code of Practice says this at §3.25:

A CHIS, whether or not wearing or carrying a surveillance device, in residential premises or a private vehicle, does not require additional authorisation to record any activity taking place inside those premises or that vehicle which takes place in his presence. This also applies to the recording of telephone conversations or other forms of communication, other than by interception, which takes place in the source’s presence. Authorisation for the use or conduct of that source may be obtained in the usual way”

The material obtained in this manner is admissible. Neither the acquisition of the content of the telephone conversation through a listening device or if recorded by a CHIS is unlawful: authorisation is purely voluntary: s 80 RIPA and C v Police and Secretary of State, IPT/03/32/H

There is no qualitative difference to the nature of the material that is the subject of the surveillance, yet the protections traverse the panoply of authorisations available through the legislative regime. This position will remain if the Bill is enacted in its current form. A similar case can be made out in respect of other investigatory techniques that would be available to the state when the Bill becomes law.

It is worth reflecting on Edward Snowden again. Whether you consider him someone who acted in the public interest or treasonously, it is inarguable that he exposed unlawful acts by governments (our own and foreign). He prompted this change in the law. International standards require it to be sufficiently clear and precise. The ISC, chaired as it is by an eminent QC, no doubt had this at the forefront of its mind when it made its recommendation. If it fails to meet the standards it will be found wanting either here or in Strasbourg. The ISC’s recommendation was the act of a responsible oversight body; in rejecting it the government has acted like a petulant child. It says the Independent Reviewer of Terrorism Legislation, David Anderson QC’s report has heavily influenced the proposed statutory framework. If it was, it has singly failed to appreciate why it was perceptively entitled A Question of Trust.

Departures from open justice in criminal proceedings: Guardian News and others v R and Incedal

On 11 November 2014 Erol Incedal was convicted of possessing a document containing information of a kind likely to be useful to a person committing or preparing an act of terrorism: s 58(1) (b) Terrorism Act 2000 (the 2000 Act). The jury had been unable to reach a verdict on a more serious charge of engaging in conduct in preparation of terrorist acts: s 5(1) Terrorism Act 2006 (the 2006 Act). Following his conviction Incedal was sentenced to 42 months imprisonment. There is nothing unique about the prosecution of these types of offences, indeed according to Home Office figures, in the year ending 31 March 2015, 50% of those charged with terrorist offences were charged with s 5 offences and during the same period there were 31 convictions (across the range of terrorist offences). What makes the Incedal case different was not that it took place almost entirely in secret – although this fact is troubling – but the role of the media in challenging the orders made in the course of the proceedings.

Following the charges Incedal (and a co-accused) were brought before the Westminster Magistrates’ Court where orders were made under s 4(2) and 11 of the Contempt of Court Act 1981 (the 1981 Act). The former provides, in summary, for the court to postpone for such period as it thinks necessary publication of any report of the proceedings where it appears necessary to avoid a substantial risk of prejudice to the administration of justice. The latter provides for a relevant court – one empowered to withhold a name or other matter – to give directions prohibiting publication as necessary for that purpose. At a later preparatory hearing before the Central Criminal Court the prosecution sought an extension of these orders and an order that all or part of the proceedings should be heard in private. This was supported by what the court of appeal called “certificates” signed by the Secretaries of State for the Home and Foreign Office. These contained schedules of supporting material. The defendant and representatives of the media were provided a redacted copy of the certificates but did not see the schedules. It was canvassed at this hearing that “certain media representatives [would be permitted to] attend some of the private part of the trial, subject to stringent undertakings”. The judge subsequently prepared two judgments (one open and one closed) and ordered that the entirety of the trial should be heard in private, the names of the defendants should be withheld and that reports of the hearing and his open judgment should not be published until after the trial or further order. The media appealed the orders pursuant to s 159(1) of the Criminal Justice Act 1988 and the Criminal Procedure Rules, Parts 6 and 40 (wrongly referred to as the “CPR” in the judgment, not Crim PR pursuant to Rule 2.3(2)(b)).

The Court of Appeal allowed the appeal in part, varying the terms of the order to a limited extent and discharging in certain respects the first order. Again the judgment was in open and closed form. In summary it held that parts of the trial could be heard in open court, that the defendants could be identified and permitted the media to report the proceedings contemporaneously. Its reason for holding that some of the proceedings could be heard in camera was that there was “a significant risk – at the very least a serious possibility – that the administration of justice would be frustrated” if the trial was held in public since the Crown may be deterred from continuing with the prosecution if the order was not granted. The Court of Appeal directed that “a small number of accredited journalists might be invited to attend the ‘bulk’ of the trial” on confidential terms and any publication would be reviewed at the end of the trial.

The practical effect of the orders was that the trial was held in three distinct stages: stage one was in open court; stage two was in camera but included accredited journalists; and stage three was in camera and excluded journalists. Notes of stage two proceedings could be taken but not removed from court and there could be no disclosure of what took place. The stages were kept under review and varied as the circumstances necessitated. Indeed at the prosecution’s behest significantly more of the trial was held in open court. At the end of the trial, the media asked that the prohibition on reporting some aspects of it be lifted but this was dismissed by the trial judge on the basis that “nothing material had changed in light of Incedal’s acquittal to justify the relaxation on the prohibitions”. This was subject to a further appeal.

The court began by dealing with a number of procedural matters. It had the power to receive evidence not provided to all parties to an appeal: s 159(5) (b) of the 1988 Act and Rule 40.5(2) and 6(f) of the Crim PR. In addition although permitting accredited journalists access to in camera proceedings was unusual, there was an “analogous course” contemplated in relation to sentence review hearings where a defendant provides assistance to law enforcement: ss 71 to 75 Serious Organised Crime and Police Act 2005 and R v P and Blackburn [2007] EWCA Crim 2290 (discussed in chapter 7 of Covert Policing Law & Practice). It then considered the principles in relation to ordering a departure from open justice in national security cases.

First was the independence of the decision-making of the Director of Public Prosecutions. This autonomy is subject only to the superintendence of the Attorney General and in extremely rare cases the courts by way of judicial review. In the context of a national security investigation, the Court of Appeal at paragraph 45, framed the position in the following way:

“Thus in a case involving national security when the police or the Security and Intelligence Services will put the evidence garnered in the course of the investigation before the DPP and set out their views as to why it might be necessary for an application to be made to the court for part of the proceedings to be held in camera, it must be for the DPP, and the DPP alone, subject to the superintendence of the Attorney General and the ultimate supervisory jurisdiction of the court in exceptional cases, to determine whether to prosecute and, if so, whether to apply to the court for part of the proceedings to be heard in camera.”

Second, where the DPP makes an application, the court proceeds on the basis that the principle of open justice is fundamental to the rule of law and democratic accountability. She must make “a very clear case”. The threshold is a high one: the leading authority remained Scott v Scott [1913] AC 417 where the House of Lords described it as requiring it to be “strictly necessary” and “that by nothing short of the exclusion of the public can justice be done”. The principle can only be departed from in unusual or exceptional circumstances.

Third, the decision must be made “in relation to the evidence in issue”. Where the reason for departing from the principle of open justice is said to be national security it is for the court alone to decide whether the threshold has been met. It decides whether the evidence or other material can be heard in public or not. In determining this the court will “pay the highest regard” to what the Secretary of State says in the certificate on national security issues providing there is an evidential basis for it. National security is an issue for the executive; determining whether there is evidence supporting the view and whether open justice should be fettered are decisions for the judiciary. The court may, though it is likely to be rare, depart from the views of a Minister as set out in a certificate. The test is one of necessity not whether the prosecution will continue with the case if the application is not successful. The earlier judgment of the Court of Appeal had considered that the risk of discontinuance of the criminal proceedings would cause serious prejudice to the administration of justice. Some support for this approach could be found in R v A [2006] EWCA Crim 4 and R v Wang Yam [2008] EWCA Crim 269.

In Incedal the Court of Appeal made it clear that the constitutional position required that the correct approach of the court was for it to consider whether the nature of the evidence made it necessary to hear the evidence in camera rather than any view expressed by the DPP. The court, at paragraph 57, held:

“Determining the matter on the basis of the DPP’s view would remove from the court its proper constitutional function of determining whether a departure from the principle of open justice would be necessitated, as the decision would rest on the implicit threat of the DPP not to prosecute unless the court were to defer to the view of the DPP on the effect of the evidence in issue rather than making its own determination of the effect of the evidence in issue. The proper approach of the court is to examine the nature of the evidence and to determine the effect of hearing it in public. Deciding the issue on the basis that the DPP might not continue with the prosecution does not satisfy the test of necessity. In effect, it transfers the decision on whether to depart from the principle of open justice to the DPP.”

If the court decides not to make an order that evidence is not heard in camera and the DPP decides to prosecute, the function of the executive is to give its full cooperation and assistance. The intelligence agencies must “abide by the rule of law”.

After the making of an order, the court remained under an obligation to subject its decision to continuous review. This, held the Court of Appeal, imposed a very substantial burden on the trial judge who would normally be assisted by trial counsel in the case. However, it also expressed the view that in some national security cases the court may be “greatly assisted by an independent lawyer assigned to provide assistance”. Such a lawyer would need to be vetted and provided to the judge in appropriate cases.

In the Incedal case, the court expressed the view that the making of an order that accredited journalists should be permitted to attend made the management of the trial difficult. In future, it “should hesitate long and hard before it makes” a similar order. As to the currency of the existing order, the court held that whilst it was not the function of the court to hoist the national security flag and the court automatically saluting it, the evidence in the case made a departure from the principles of open justice strictly necessary. The appeal was therefore dismissed. By way of finale, the court expressed understandable concern that closed judgments were not retained on the file or within the court building. It was not clear where they were located. The court ought to be able to consult them in the interests of consistency of approach. The court directed the Registrar to establish a working group to advise on the course of action it should take in future cases.

This is an important constitutional decision, underscoring the need for the office of the Director of Public Prosecutions to be independent both in name and in practice. Lines of demarcation were necessary between the Crown Prosecution Service and intelligence agencies – the agencies could not revoke its cooperation if the Crown decided to prosecute after a failed application for proceedings to be held in camera. It also touched on the question of Ministerial certificates: these are creatures of the executive and subject to the duty of candour (discussed here). However, the limitations on oversight of the application of this doctrine as far as the agencies are concerned are limited in scope and in theory open to abuse. Complaint lies to the Investigatory Powers Tribunal and there have been no reported decisions on the issue.

There are a number of key practical points arising out of this important judgment:

  1. An application the 1981 Act must be determined by reference to the evidence in issue, not the risk of discontinuance
  2. The DPP must be scrupulously independent of the police and intelligence agencies
  3. Orders under the 1981 Act providing for access by accredited journalists to some or all of in camera proceedings are unlikely to be made having regard to the burden this places on the judge in case management terms
  4. Where an order is refused and the DPP decides to prosecute, the police and agencies must cooperate with the prosecution
  5. In obtaining Ministerial certificates the agencies and executive are under a duty of candour.

© Simon McKay (2016). Please ask for my permission before using or copying the material contained in this blog. If you do, attribute it to me, with a link to the original content. This is opinion, not legal advice.

Of Prime Ministers and Presidents: a review of the ISC Report on the Investigatory Powers Bill, actually

The Intelligence and Security Committee (ISC) has published its report on the Investigatory Powers Bill (the Bill). The scope of the report is limited to the work of the intelligence agencies: the Joint Committee will examine all other aspects of the Bill and is due to publish its findings later this week.

The report does not get off to a good start for the government, making the early and well-made observation that it fails to deal with the entirety of the intelligence agencies armoury: “[the Bill] is handicapped from the outset in terms of the extent to which it can provide a clear and comprehensive legal framework to govern the use and oversight of investigatory powers”. It is “a significant missed opportunity”; an observation it is difficult to demur from.

But if the Home Office hoped the report was getting the bad news out of the way first, these were dashed. In paragraph 5, the hard-hitting report expresses three important concerns:

“i. Overall, the privacy protections are inconsistent and in our view need strengthening. We recommend that an additional Part be included in the new legislation to provide universal privacy protections, not just those that apply to sensitive professions.

  1. The provisions in relation to three of the key Agency capabilities – Equipment Interference, Bulk Personal Datasets and Communications Data – are too broad and lack sufficient clarity.

iii. In addition to these major issues of principle, there are a number of more detailed matters requiring specific amendments. These can be found at the end of this Report.”

This was remarkable – reminiscent of Hugh Grant’s prime-ministerial put down of a presidential Billy Bob Thornton in the movie Love Actually and universally welcomed by lawyers and campaigners. Buzz feed’s @jamesrbuk described it as a “savage report” at BBC security correspondent @gordoncorera “very critical” and leading surveillance law QC @rydermc strongly critical. Capturing the mood of the interested element of the nation, @3i5 (Eric King) mused that the new ISC chairman and former Attorney General Dominic Grieve QC would be off the Home Secretary’s Christmas card list this year.

The theme of the report was palpable frustration – the Bill “failed to provide the clarity and assurance which is badly needed” – and talked of rebuilding the public’s confidence in the work of the intelligence agencies. This is a really important point – it is significant that terror watchdog, David Anderson QC’s excellent report on surveillance powers, was entitled “A Question of Trust” – there is little of this around when it comes to our intelligence agencies at present. In a civilised society intelligence agencies can only legitimately exist where there is accountability and public confidence. Missing this fundamental principle in the drafting of the Bill is a serious error, giving the impression of it being seen as an opportunity to make compliant that found wanting either through leaks or adverse judgments and not seeking to give some integrity to the activities of our intelligence agencies. The ISC has mandated the government to address its concerns and publish a new draft.

Refreshingly, the ISC has stated in unequivocal terms that consistency is needed in terms of privacy protection. It is a remarkable feature of the existing regime that to intercept a telephone call, for example, a ministerial warrant is required and that the existence of the warrant or disclosure of the product without lawful authority is a criminal offence. If the same telephone calls are recorded on a listening device placed in a car using a hands free facility, an intrusive surveillance authority is needed, authorised internally and approved by a Surveillance Commissioner. If a covert human intelligence source hears the conversations only an internal authorisation is necessary. Incidentally the latter two resources do not need to be authorised: s 80 of RIPA, discussed here, makes clear such activities are not made unlawful in the absence of authority. The same content – telephone conversations – qualitatively the same interference with privacy but wildly different protections. A new part should be added to the Bill says the ISC to deal with this.

The ISC also expressed concerns about what is now called equipment interference – computer based covert investigations – and that there was a need for the Bill to provide a single legislative regime for this. Further repeal of the Intelligence Services Act 1994 (which placed GCHQ and MI6 on a statutory footing) is recommended.

On the subject of bulk interference, the ISC has stated bluntly that it has not been provided with, “sufficiently compelling evidence as to why the Agencies require Bulk Equipment Interference warrants, given how broadly Targeted Equipment Interference warrants can be drawn”. Relying on the evidence of GCHQ itself that bulk interference was an exact science, the ISC has advised that the proposal for bulk equipment interference warrants are removed from the Bill. It is not just bulk equipment interference that should in the Committee’s opinion be consigned to the scrapheap, so too, “class bulk personal dataset warrants” (large datasets containing personal information about a wide range of people).

On the question of communications data, the ISC described the current proposed legislative provisions as “inconsistent and largely incomprehensible”. Importantly, the basis for their acquisition must specified in the legislation, “it is not sufficient” said the ISC, to rely on that which has come to the agencies rescue in some of the legal challenges, “internal policies or Codes of Practice”.

Other criticisms are made, including:

  1. The conflation of national security and economic wellbeing of the United Kingdom where relevant to national security (also the subject of criticism in Covert Policing Law & Practice).
  2. The absence of any specifity about the term “operational purpose”.
  3. Confusion over “Targeted Interception” and “Targeted Examination” warrants.
  4. The absence of effective review of timescales where material has been obtained without a warrant or on an urgent basis
  5. Absence of the power to refer certain matters to the ISC for investigation

If this was a bad day for government covert surveillance policy, it was great day for those seeking to strike the correct balance between the use of secret powers and accountability. In the end government failed to make its case and fell short; unable to justify a compelling case for the use of invasive and intrusive secret resources that have scope for huge abuse with little accountability. It is a bit like the scene from Samuel Beckett’s Waiting for Godot when Vladimir says to Estragon, “I don’t understand” and he replies, “use your intelligence, can’t you?” (Vladimir then uses his intelligence). Finally, says Vladimir, “I remain in the dark”. As Dominic Grieve said in an interview earlier today on the World at One, “when we asked why such a power was being provided, we didn’t necessarily get a clear answer”.

© Simon McKay (2016). Please ask for my permission before using or copying the material contained in this blog. If you do, attribute it to me, with a link to the original content. This is opinion, not legal advice.

Recruiting informants and radicalisation: legal issues

Stories that young men have been radicalised following attempts by the Security Service (MI5) to recruit them as informants (technically known as Covert Human Intelligence Sources, or as the former Chief Surveillance Commissioner Sir Andrew Legatt observed, “inelegantly” CHIS) appear with increasing regularity. Last year there were reports that Mohammed Emwazi, “Jihadi John”, may have been turned towards Islamic extremism following a failed pitch by MI5 for him to work for them. Last month a broadly similar account appeared in relation to Siddhartha Dhar, Jihadi John’s apparent successor. Michael Adebowale, one of Lee Rigby’s killers, claimed that he too was the subject of a failed recruitment bid by the agency. In its evidence to the Intelligence and Security Committee MI5 neither confirmed nor denied (NCND is discussed here) whether this was true on the grounds that it would damage national security to do so. What MI5 was prepared to say was that it had not carried out an assessment of his mental health [see paragraphs 126 to 128 of the report].

The issue of radicalisation following a failed attempt to recruit is one deserving of proper research but this paper examines the legal issues arising. The Chief Surveillance Commissioner, who does not have responsibility for the security services, identified a similar issue in his latest report:

“…less common, but still encountered, the failure to authorise a CHIS promptly as soon as they have met the criteria; and in many cases (more typically within the non-law enforcement agencies) a failure to recognise or be alive to the possibility that someone may have met those criteria”

This encapsulates both the problem and the subject of this blog: defining CHIS and evaluating the considerations on attempting recruitment.

The definition is found in the Regulation of Investigatory Powers Act 2000. It is often presumed in s 26(8). This provides:

“(8)    For the purposes of this Part a person is a covert human intelligence source if—

(a) he establishes or maintains a personal or other relationship with a person for the covert purpose of facilitating the doing of anything falling within paragraph (b) or (c);

(b) he covertly uses such a relationship to obtain information or to provide access to any information to another person; or

(c) he covertly discloses information obtained by the use of such a relationship, or as a consequence of the existence of such a relationship”

There is some elaboration of the terms in s 26(9):


(b) a purpose is covert, in relation to the establishment or maintenance of a personal or other relationship, if and only if the relationship is conducted in a manner that is calculated to ensure that one of the parties to the relationship is unaware of the purpose; and

(c) a relationship is used covertly, and information obtained as mentioned in subsection (8)(c) is disclosed covertly, if and only if it is used or, as the case may be, disclosed in a manner that is calculated to ensure that one of the parties to the relationship is unaware of the use or disclosure in question”

These provisions are explained in Covert Policing Law & Practice:

“Attempting to give the definition some clarity is not easy but the relationship is essentially tripartite—the ‘authority’ one part, the source the other and the ‘target’ the third. Subsection 9(a) requires the source to establish or maintain ‘a personal or other relationship with a person’ for the purposes of the two subsections that follow”.

However s 26(7), the subsection immediately prior to the substantive definition, expands this significantly. It provides:

“7.     In this Part—

(a) references to the conduct of a covert human intelligence source are references to any conduct of such a source which falls within any of paragraphs (a) to (c) of subsection (8), or is incidental to anything falling within any of those paragraphs; and

(b)references to the use of a covert human intelligence source are references to inducing, asking or assisting a person to engage in the conduct of such a source, or to obtain information by means of the conduct of such a source”

So, it can be seen, that any act incidental or preparatory to the proposed recruitment of a CHIS falls within the definition and, to be compliant, requires authorisation.

The significance of authorisation is that s 29(5) creates an obligation to appoint someone with responsibility for the welfare and security of the source as well as a quite different person to oversee the use made of the source. The Covert Human Intelligence Sources Code of Practice on sets out the practicalities of this at paragraph 6.14:

“Before authorising the use or conduct of a CHIS, the authorising officer should ensure that a risk assessment is carried out to determine the risk to the CHIS of any tasking and the likely consequences should the role of the CHIS become known. The ongoing security and welfare of the CHIS, after the cancellation of the authorisation, should also be considered at the outset” (emphasis added).

There is no evidence the ISC considered the provisions of RIPA 2000 during its examination of the Lee Rigby killing. It is not known whether Adebowale was the subject of an attempted recruitment or not but carrying out an assessment of Adebowale’s mental health may have been a requirement of any risk assessment if he had been.

What, if any conclusions can be reached from this short analysis? There is no question, that recruitment of agents is an invaluable resource for the agencies and law enforcement. Sun-tzu’s in The Art of War, described informants as ‘the Lord’s Treasure’ and his five categories of ‘spies’ included local and double agents. Their importance in the fight against terrorism and crime is unassailable. However, an attempted recruitment carries with it the obligations of authorisation, which in turn, triggers the need for effective risk assessment and oversight. The OSC in the context of non-agency public authorities was concerned “about the huge variation in the standard of risk assessments” and that there is an effective “firewall” between those with responsibility for security and welfare and exploitation of the source. The Adebowale case may point towards similar problems within the intelligence community. One further conclusion may also be possible and that is that there was a fundamental failure by the ISC to examine what must have been one of the critical questions the murder of Lee Rigby gave rise to.

© Simon McKay (2016). Please ask for my permission before using or copying the material contained in this blog. If you do, attribute it to me, with a link to the original content. This is opinion, not legal advice.

The Office of Surveillance Commissioner’s Report 2014 – 2015: a review

Sir Christopher Rose, former Chief Surveillance Commissioner, has published his last report. Those who have read his annual reports may think a fair epitaph of his tenure might be, to quote `Shakespeare, “no legacy is so rich as honesty”. Others could fairly observe that his reports have become increasingly more acerbic. This last report, attempting to find some neutrality, certainly holds no punches.

The oversight of undercover officers, or “relevant sources”, has “had a considerable impact” in terms of volume of work. This may be counter-balanced by the stark reduction in directed surveillance authorisations from the Department of Work and Pensions. The Commissioner hypothesises reasons for this – greater use of overt investigations, collaborative working etc. What Sir Christopher does not consider is whether it might have something to do with an interpretation of s 110(2) of the Social Security Administration Act 1992, which may provide a legislative basis for an inspector to conduct some form of covert surveillance in connection with inquiries into benefit fraud. The Department could be relying on the Code of Practice on Covert Surveillance and Property Interference, at paragraph 1.15:

“Similarly, an authorisation under the 2000 Act is not required if a public authority has another clear legal basis for conducting covert surveillance likely to result in the obtaining of private information about a person. For example the Police and Criminal Evidence Act 1984 provides a legal basis for the police covertly to record images of a suspect for the purposes of identification and obtaining certain evidence”.

There has been no ruling on the applicability of s 110(2) but its scope may have given the Department of Work and Pensions some confidence, though it may be misplaced.

It is now clear that in respect of undercover officers there will be some limited reporting. The number of authorisations notified, cancelled and applications for prior approval will be reported. In this reporting year there were 1095 notifications, 770 cancellations and 46 submitted for prior approval. The OSC emphasise these figures need careful interpretation: a single covert operation may deploy more than one undercover officer, for example.

Overall the use of covert policing resources overseen by the OSC is down and in the case of property interference and directed surveillance significantly down (by 598 and 1331 respectively). The use of CHIS by law enforcement is down, although there is a slight increase of use by non-law enforcement public authorities.

Irregularities were up slightly – there was nothing unusual about their nature – pre-emptive surveillance activity before the authorisation is in place, leaving listening devices on after cancellation and failing to authorise use and conduct. Sir Christopher identified no evidence that the irregularities were intentional or motivated by bad faith. Importantly, the Chief Commissioner recorded that “failure to obtain an authorisation under the Acts for which I have oversight is not unlawful”. This is by virtue of the operation of s 80 RIPA, dealt with here.

Unusually this year’s report identifies instances of misconduct involving non-compliance with RIPA. These were: (i) use of a commercial surveillance device; (ii) the deliberate extension of the parameters of an existing authority which resulted in the acquisition of material subject to legal professional privilege; and (iii) the fabrication of reviews of use and conduct authorities. The consequences have been catastrophic for the officers concerned. In Covert Policing Law & Practice the propriety of disciplining officers for “breaching” RIPA is questioned. There is nothing unlawful in undertaking surveillance activities without an authorisation. Much will turn on the nature of the charges but caution should be exercised.

Inspections in respect of relevant sources will not result in the “complete eradication of rogue, unregulated activity by an undercover officer [and] compliance errors are inevitable”. However Sir Christopher is of the view that the chances of this are significantly reduced. In respect of authorisations during the reporting period, two issues arose. First, the completion of risk assessments could be improved upon – many are formulaic and contain little of relevance about the risks to the officer based on the operational scenario. Second, there is no effective de-brief on cancellation. This is constant theme of the OSC’s reports albeit in respect of the range of Part II activity. Meaningful appraisal of a deployment is essential and it will be a core aspect of future annual inspections.

Sir Christopher has had little positive to say about the Protection of Freedoms Act 2012 in previous reports. In this, his final report, his conclusion is that “there is little evidence that” the introduction of judicial approval for local authorities carrying out directed surveillance “was a move for the good”. This is interesting, not least because the Investigatory Powers Bill essentially “rolls out” a scheme of prior judicial authorisation for a range of covert surveillance activities. He also laments the lack of training for magistrates.

A series of issues are identified where public authorities need to improve. These include:

– poor analysis of intelligence cases

– the use of formulaic language in relation to, in particular, the assessment of risk of collateral intrusion

– lack of effective assessment of proportionality

– cutting and pasting

– ineffective de-briefing on cancellation, including handling of product and recovery of technical equipment

– over reliance by AO’s on pre-prepared entries on the application form and a failure to consider properly necessity and proportionality

– the failure to authorise a CHIS promptly

– poor risk assessments,

– effective discharge of the duties and compliance with the requirements of s 29

– a lack of training

Curiously, Sir Christopher ends the narrative of his report by referring to public reassurance. He asserts “the public can be reassured that [Part II] powers are almost always used only when necessary and proportionate”. The strength of this statement is undermined in two ways. First, his suggestion that the media is misreporting improper surveillance activities, “we also continue to see media articles about the use of surveillance and other covert activities that could lead the public to assume that “Big Brother is alive and well”. This is a curious assessment of the present climate where the role of the press has been central to exposing systemic impropriety, indeed illegality on the part of the state. Second, the efficacy of the oversight scheme depends largely on what Commissioner’s are told. This weakness has been exposed in Chatwani, discussed here. It will inevitably lead to ever increasing calls for the introduction of judicial authorisation.

© Simon McKay (2016). Please ask for my permission before using or copying the material contained in this blog. If you do, attribute it to me, with a link to the original content. This is opinion, not legal advice.