Transparency data

Justice practitioners’ roundtable: 12 March 2024 (accessible version)

Updated 20 August 2024

Attendees:

Chair: Jonathan Fisher KC

Host: Fiona Rutherford – Chief Executive, JUSTICE

Patrick Rappo – Partner, Reed Smith LLP

Anand Doobay – Partner, Boutique Law LLP

Allison Clare KC – Barrister, Red Lion Chambers

Leila Gaafar – Barrister, 6KBW College Hill

Timothy Folaranmi – Managing Associate, Barrister, Mishcon de Reya LLP

Sam Brown – Senior Legal Advisor, Jersey Law Officers’ Department

Alex Davidson – Barrister, 2 Bedford Row

Ross Dixon – Partner, Hickman & Rose LLP

Faye Rolfe – Barrister, Red Lion Chambers

Charlotte Atherton – Barrister, Exchange Chambers

Anita Clifford – Barrister, Red Lion Chambers

Home Office officials

1. It was highlighted that a number of findings have emerged from stakeholder meetings undertaken thus far. First, all stakeholders had been clear that the CPIA’s framework was broadly acceptable, with no significant statutory changes required. Second, on both sides – prosecution and defence – there was a wish expressed for early engagement with the disclosure process. The question was as to precisely how early: pre-charge appeared to be less popular than immediately after transfer to the Crown Court. It became more complex to ascertain precisely what this would look like, particularly in relation to different types of cases, whether serious fraud, RASSO^{[footnote 1]} cases and the more general run of criminal cases before the courts. If there were early disclosure hearings, for example, there was a question of how that would work in practice, in particular where the defence team did not, or did not want to, engage. This raised the question of sanction.

2. There was some support for the ‘keys to the warehouse’ approach, however, the majority on both sides agreed that there was no hurry to go down that road. The defence would, especially in legal aid cases, struggle to deal with it. Issues of data protection and redaction also arose.

3. The Review had been told that there are real problems in the magistrates’ courts because disclosure is not happening as envisaged. Thought was given to how this might be dealt with and to consider what is going on with respect to training, particularly of police officers. There was a general awareness of artificial intelligence (AI), and an understanding that it was not a silver bullet albeit may be of some use.

The CPIA

4. Participants were asked to comment on whether they considered the CPIA fit for purpose. One participant observed that the CPIA placed the burden for disclosure on the prosecution whereas it was primarily the defence who wished to see the material. It was suggested that gathering and recording material should remain with the prosecution, but determining relevance is conceptually beyond the prosecution. When considering Data Protection Act 2018 (‘DPA’) issues, it was noted that these would generally arise before an investigation begins. The participant felt that disclosure was too much of a burden to place on the prosecution, and therefore favoured a ‘keys to the warehouse’ approach so long as there was the ability to redact as necessary. Furthermore, it was suggested that the CPIA is overly onerous and prone to problems, which has been exacerbated by large volumes of data and lack of proper AI and/or review mechanisms by either side. It was agreed that whilst the CPIA worked and was fit for purpose, the question is about its application in practice. Another issue that was raised was that, even with defence engagement and the service of detailed defence statements, no update by way of disclosure was necessarily provided by the prosecution.

5. Discussion turned to SFO cases that had failed in recent years. It was suggested that, in each case, the problem was not the volume of material. Rather, disclosure failures had arisen after the SFO had identified and reviewed the relevant material but had taken the decision not to disclose it. It was therefore suggested that something was going wrong with how the CPIA was being applied in those cases. It was suggested that one solution might be the ‘keys to the warehouse’ approach, but there are also practical problems with this, particularly in relation to defendants who were not well-funded. Legal aid firms were more concerned about the‘keys to the warehouse’ approach because they would not be in a position to review the relevant material.

Block listing

6. There was discussion regarding the use of ‘block listing’ in high volume cases, as a way of relieving some of the scheduling burden on the prosecution. Participants considered how this provision could be used effectively to flag to the prosecution and defence material which met the disclosure test but reduce the require to write a description for all relevant items. It was noted that this approach is used effectively in some live cases, however there remains a dissonance between and within the AG’s Guidelines and Code of Practice about what precisely is permissible.

7. One participant reported that there appears to be a broad agreement amongst financial professionals that, in fraud cases, the defence should be entitled to their own documents. This includes their own emails, outlook calendar and other documents to which a defendant would have had authorised access to during the indictment period. It was noted that free access to these documents worked in fraud cases because the defence would know what to look for, rather than being overwhelmed by the volume of material. It was also acknowledged that the defence usually end up getting this material late in proceedings, when the prosecution have done the work, which does not necessarily place the defence in a good position. Participants suggested that this could perhaps be changed by means of amending the CPIA Code, rather than the CPIA itself. Some participants observed that the downside to the CPIA is that sometimes prosecutors can see the force in the argument for disclosing certain material but cannot justify it by reference to the disclosure test, and there was a fear of setting a precedent. There was a general agreement that the Attorney General’s Guidelines on Disclosure sets out an effective presumption in favour of disclosure in relation to certain categories of material. Such disclosure could be achieved very early in the process.

8. On that theme, the view was expressed that the ‘keys to the warehouse’ approach would not be feasible in relation to third party or prosecution material because the defence would not know what it was looking for as it did with its own material. There were also issues in relation to cross disclosure and multi-handed cases. This should be directed in relation to certain categories of material to allow a level playing field. Another participant noted that, sometimes, one defendant will have looked ahead and copied their emails which created unfairness between defendants. Stricter provisions could be put in place where GDPR issues arose.

Data protection

9. Discussion turned to data protection. It was felt that there is generally an issue with over redaction of documents. Whilst caution is understandable, redaction takes up a huge amount of time and is not always required to the extent that it is done. It was observed that serious fraud cases normally concern allegations of offences committed in the course of a defendant’s employment as a CEO, for example. There is therefore a need to be able to look back at what the defendant may rightly have had access to.

10. Whilst it was suggested that this analysis applied only to fraud cases, consideration was also given to how it might apply to other cases which might involve large volumes of digital material, such as modern slavery cases. It was proposed that there is sometimes a presumption in favour of disclosure in relation to certain categories of material, and/or certain categories of case. One example might be where a person was charged in the context of their employment. Generally, the issue was achieving access to devices other than the defendant’s own; ordinarily the defendant would have access to their own phone.

11. Another participant observed that in relation to the defendant’s own material no data protection issues should arise, whereas others took a different view, noting that complications arose where a defendant’s material contained criminal material. Some defendants were still working for the same employer at the time of the prosecution which caused inherent unfairness.

Digital materials – reverse burden of reviewing material?

12. There was a question as to whether a separate category did or should exist for digital materials. The CPIA rests on the supposition that in investigating, the prosecution have read everything in its possession. This is not the case in relation to digital material, which is simply collected, and the investigation team makes its own decision about what should be reviewed and how to review it. It was suggested that there should perhaps be a separate category for digital material where the position is reversed in terms of who bears the burden of review. Issues about data protection could then be addressed at the other end of the process – there is sufficient protection in the CPIA to allow the material to be shared with the defence team. The material can then be checked for data protection issues before anything is deployed at trial.

13. It was observed that this approach meant that that the prosecution do not get ambushed. The burden did not fall on the prosecution and in a way it could not do so, because the prosecution are only reviewing material in order to comply with the CPIA. Some noted that review and scheduling is not helpful for digital material at present. Furthermore, it is thought that this approach can significantly reduce the volume of unused material and that any dispute around search ranges can be resolved. However, if the search results still produced is too much, this would become a case management issue.

14. It was pointed out that where a defendant is not at the top of the indictment, they may not have visibility of the issues being explored by the defendants who are. Such a defendant would have no way to understand what issues to look for when reviewing the material. It was pointed out that many solutions that had been discussed assumed that the defendant knows what is in the material when that is not necessarily the case.

15. It was then noted that there would be some defence teams who may find it difficult to manage disclosure obligations in this way, namely those who are publicly funded and lack the resources, whether financial or technical, to be able to engage with the analysis of material as a privately funded team might. A privately funded defendant may want full access to emails and be represented by a firm that is able to look at that material when required. The position is likely to be different for a client in custody represented by a legal aid firm of solicitors and a legal aid barrister. Participants were of the view that there is a risk of creating a two-tier system if too much onus is placed on the defence to carry out the review of digital material. It then becomes a question of who can afford to review potentially exculpatory material.

The position in Jersey

16. It was noted by one participant that defence statements have not been a feature of criminal proceedings in Jersey until relatively recently. Historically, there has been a reluctance to plead defences at all and the defence are not encouraged to do so. The statutory requirement to file a defence statement was expressly resisted at debate stage by the president of the Jersey Law Society, relying on pre-1996 cases. Disclosure is therefore now a prosecution led process, and there was little engagement from the defence for the reasons identified and because most of the experience came from English-trained lawyers within the Law Officers’ Department. Often there are no requests at all for disclosure, or very many which are irrelevant, misconceived or seeking the keys to the warehouse. In Jersey, there are no section 8 provisions, so there is an enormous amount of pressure on individual prosecutors to second guess defences which may or may not arise, meaning the position is almost pre-1996. So far, there have been no major problems and the courts have been quick to intervene to manage any pre-trial disclosure difficulties. It was recognised that similar themes are present, but there are struggles with a culture that has not previously encouraged defence pleadings at all.

Non-engagement and sanctions

17. Participants were also asked to consider how the courts might deal with a situation where the defence does not engage with an early hearing to discuss issues in the case and identify reasonable lines of inquiry. One suggestion was to simply give the defence their own material and carry on with the proceedings. Another was for the judge to order section 8 applications by a particular date; although not all participants were persuaded that this would work in practice, some were more optimistic that it would if a clear structure for it was set up. It was observed that there could also be lack of engagement on the part of the prosecution. Therefore, there ought to be a relevant sanction on the prosecution. This might take the form of financial consequences.

Early engagement

18. One potential downside to early engagement was suggested: namely that if the defence do not have access to the material for a very long time then it may be difficult to identify the issues in the case at at early stage. It is important to recognise that serious fraud cases take time and placing too much pressure could create its own problems, e.g. by fixing a trial date too soon. Under the current process, longer timelines are arguably more realistic, whereas trying to push disclosure into the early part of the timeline is not. Others noted that, where a defence team is not ready to identify the issues early in proceedings, this can be explained to the court However, there may be other cases where the position is quite different. Another observation participants made is that early engagement should avoid disclosure issues being raised close to trial; this should include section 8 applications.

19. It was thought by some participants that if disclosure management documents are completed at an early stage, with early judicial and defence engagement, the system could work much more effectively. This would only work, however, if there was an obligation to produce the document and a corresponding obligation for the defence to respond to it. Furthermore, it was suggested that there ought to be a uniform standard for disclosure management documents across the board, otherwise their utility would be limited.

AI

20. Some participants had used AI and found it to be a useful tool. It was noted that private defence firms are already using advance technology, with AI functions, to perform material review tasks. Some programmes were not necessarily good for reviewing, but were more effective for scheduling purposes. Questions were raised as to whether the summaries produced by AI for schedules were sufficiently nuanced. AI could refocus the algorithms when the issues changed, which could not be achieved manually. The considerable fees associated with the use of such programmes was acknowledged. Centrally hosted databases that defendants could access, with suitable firewalls in place, could help to address funding issues.

21. At present, AI is not being fully utilised owing to wariness and the lack of judicial stamp of approval. Participants felt that it worked if the prosecution shared the data set. Some participants had been doing the same but in reverse, identifying processes that would satisfy the section 2 notice^{[footnote 2]}. It was proposed that the defence were given the metadata as a list, even if there was block listing of material. Metadata was thought to be potentially very helpful generally.

1. Rape and serious sexual offences ↩

2. Section 2 of the Criminal Justice Act 1987 confers investigatory powers on the SFO. These include the power to require a person to answer questions (or otherwise furnish information) or to produce documents. ↩