The Mercedes-Benz NOx Emissions Group Litigation Aurora Cavallari & Ors v Mercedes-Benz Group AG & Ors

This is an application by the Claimants to vary an order dated 19 January 2024 made by Mrs Justice Cockerill, together with a consequential variation at paragraph 37 of the March order made by both Mrs Justice Cockerill and myself following the March CMC. The January order provided:

The March order, which was made following selection of samples from the Mercedes GLO, provided:

The Claimants rely on two statements from Mr Lawson, and the Defendants on one by Mr Bennett and correspondence to which I have been taken. I have already ordered that a redacted version of the statement from Mr Lawson be provided excising reference to the confidential parts of the A2L files cited and annexed to that statement.

I am grateful to Mr Campbell for the Claimants and Mr Purnell for the Mercedes Defendants for tailoring their submissions so that this hearing could take place in public, which is, as is well known, of considerable importance.

An A2L file is a file relating to the firmware within a vehicle's software which is a human and machine-readable text file that serves as a map or guide to the firmware, indicating which sections of it reflect what code and what functions that code controls. It is common ground that there will be parts of the firmware which are not relevant to the emissions control system or the existence of prohibited defeat devices. There is a debate about how large a part that might be expected to be.

The firmware will also, it is not disputed, contain confidential information owned by the Mercedes defendants.

The January and March Orders envisaged, as I have set out, that the Mercedes Defendants would be entitled to redact/extract irrelevant material. The January Order envisaged that the parties would agree relevant parameters in relation to the redactions.

There was a debate between the parties in correspondence as to the relevant parameters as required by the Order. That debate effectively concluded with the Defendants proposing disclosure in two tranches: the first consisting of the parts of the A2L files which relate to functions which, on the Defendants' reading, are specifically alleged in the Generic Particulars of Claim and/or by the KBA, to be PDDs; the second consisting of disclosure of the parts of the A2L files which relate to any remaining functionalities which modulate the emissions control system.

Despite some reservations on both sides, this two-stage approach was the agreed way forward by letter of 29 February 2024. That letter is said by Mr Purnell in his skeleton argument to constitute the agreement reached on the relevance parameters as required by the January Order, based upon the evidence of Mr Bennett at paragraphs 11(a) to (d). 81 files have been disclosed on the first basis relating to the 150 firmware files. It can be seen from this that an A2L file might apply to more than one firmware file. Only one A2L file has been disclosed on the second basis, and that one does not relate to any of the 150 required pursuant to the Order. It relates instead to a sample vehicle the Mercedes Defendants have proposed.

The Defendant now says that the second stage is not necessary for compliance with the January or March Orders. I disagree. The only sensible reading of the correspondence, and indeed of the evidence from Mr Bennett at paragraph 11, and of Mr Purnell's skeleton, is that the two-stage process agreed at the end of February 2024 is that required to comply with the January Order. This conclusion is strongly supported by the fact that the March Order was not drafted in such a way as to make clear – if that was really Mercedes' position at the time - that only the first agreed stage was required for compliance.

It follows from this that it is not sensibly arguable, in my view, that the Mercedes Defendants have, as they stand before the Court today, complied with the January or March Orders. It also follows that there is a need to apply for a variation of the order if they are to be permitted only to provide A2L files in relation to the sample vehicles. That application was made orally today by Mr Purnell.

The central area of controversy relating to redactions is the excision of the description strings, function names and chapter headings, along with the data from more than what is said to be around 95% of the A2L files in the first tranche. The Claimants' experts say that this makes A2L files effectively unusable, as they say they cannot discern what functions the elements perform without the descriptions. They also say that the description strings, function names and chapter headings are not themselves confidential, although they accept that the parameters and other data is, or at least may be.

Mr Campbell also argues that the claimants have identified, by reference to files found within the public domain, particular instances of specific redactions regarding functionalities which they say should not have been redacted. This, they argue, gives them concerns about what else may have been wrongly redacted.

Finally, the Claimants say that they had no proper visibility of what has been removed.

The Defendants contend that the percentage itself (95%) is irrelevant in circumstances where it is unsurprising that a large proportion of firmware will be irrelevant to matters in this litigation.

In relation to the description strings and headings, et cetera, the Mercedes Defendants do not accept that they are necessary in order to utilise the A2L. They also point to the fact that, in the past, the Claimants have said that they do not need the A2L files at all. It is said that the file can still be used without chapters and headings, using automated calibration tools readily available on the Internet. The Mercedes Defendants are also concerned that provision of this information may in fact lead to confusion on the part of the Claimants or their experts.

The Mercedes Defendants, finally, say that the disclosure was ordered in the first place initially in the context and for the purpose of choosing samples, which they point out has now been done. They therefore contend that it is not now necessary to disclose the number originally ordered, and, as I have described above, now orally seek a variation of the March Order in that respect. The Claimants counter this by saying that the files are still required to test the question of homogeneity and to help establish the representative nature of the samples selected.

Both sides have referred to the opposing sides’ previous statements made in relation to the need for A2L files. I am not assisted significantly by the fact that, on my reading of the correspondence and the transcripts, each party has, or at least may have, changed their position somewhat. It is perhaps understandable that a party's understanding of the importance of a particular document may change as the technical investigation continues. The bottom line is, and the starting point is, that these documents have already been ordered to be disclosed.

In terms of approach, I accept that the starting pointing is that a court will normally be satisfied by a statement from a solicitor with responsibility for the disclosure process that the redaction in question has been properly made. I note, however, that Mr Bennett, in his witness statement, does not state who and in what precise circumstances the redaction process was carried out. As noted at page 208 of Matthews and Malek on Disclosure, where there has been heavy redaction of documents, the court may adopt greater scrutiny to ensure that the right to redact is not being abused or too liberally interpreted, recognising that the burden is on the advocate to make out a case for inspection.

Matthews and Malek go on to identify that where there is a doubt, a court has various options. First, it may direct a party to reconsider redactions in light of the court's finding. This is effectively what the Mercedes Defendants seeks if I am against them in principle. The second is to require further explanation of what is being redacted and why. Third, the court may direct that the documents be provided unredacted on a confidential basis. It is noted in the text that this is only appropriate where the ground for redaction is the confidential nature plus irrelevance, and not appropriate for privilege. The redactions at issue before me are indeed about irrelevance, not privilege, and this is essentially the option sought by the Claimants. The fourth option is the court looking at the documents and deciding, which, it seems to me, at least at the moment, is not a practical option, where relevance may turn on a highly technical question.

Turning to the parties' arguments, the point made by the Mercedes Defendants about the percentage, at least taken alone, has some merit. There is little I can derive by way of evaluation of the appropriateness of approach by looking at the percentage alone.

I am, however, persuaded that the description strings, function names and chapter headings, as set out in the witness evidence of Mr Lawson and by reference to the examples given in the evidence and in appendix 2, that these should not be applied as irrelevant or confidential for the purposes of disclosure. This is so both for those elements accepted as relevant and those that are not, because it is difficult, it seems to me, to see why they are, of themselves and without the parameters, confidential.

Even if I am wrong about this and they are to be treated as confidential, they provide appropriate transparency as to the basis of redacting the data and the parameters, and I consider that, given that these documents are being disclosed into a confidentiality ring, there is no prejudice in determining the question of relevance or potential relevance or lack of confidentiality in the Claimants' favour at this stage. In this context, I note that, whilst ensuring particular additional safeguards are in place with regard to access and disclosing information, the Ford Defendants have not sought to redact any information from the A2L files.

In relation to the specification functionalities which were redacted, as set out in Mr Lawson's witness statement, I consider that these items neatly demonstrate the dangers of taking an overly restrictive view on relevance in the context of redactions. On the basis of the evidence before me, it seems more likely than not that the elements redacted are relevant functions which could properly form part of the Claimants' expert investigations of the issues in this case. I conclude this bearing in mind the generic nature of the pleadings at this stage, and that the absence of pleaded specificity in relation to PDDs is unsurprising. Indeed, the order made at the CMC is that the production of the specific particulars of claim is to follow disclosure, which acknowledges the fact that, by reason of the asymmetry of information, disclosure will inform the investigations prior to particularised pleading.

I do make clear, however, that I am not able to conclude that relevance was so obvious that the only conclusion open to me is that there was a deliberate redaction of material which was known by the redacting party to be relevant. Nevertheless, it is my view, for the reasons I have given, that these functionalities should probably not have been redacted.

There is, I should add, always likely to be a penumbra where both parties may have prima facie legitimate views about relevance. However, where one party's expert gives evidence, directly or indirectly, to the court that particular information is relevant to a matter being investigated within the parameters of the pleaded case, providing that evidence is prima facie rational and comprehensible, it will be given considerable weight by the court, even if the other party's expert offers a view that the information is not relevant. Ultimately, the question of relevance, in a case relating to highly technical issues, is a matter which the court is probably not going to be able to determine summarily on disclosure disputes, but, providing disclosure remains proportionate, the court will invariably err on the side of requiring disclosure to be given, providing that any appropriate confidentiality-related concerns are accommodated.

The Claimant now seeks completely unredacted files. The Mercedes Defendants seek, if I am against it on scope of redaction -- which, as I have stated, I am -- a further review process. In my view, all of the description strings, function names and chapter headings ought to be unredacted. Much of the data – which is where the confidentiality principally resides – may be irrelevant. Therefore, in the first instance, it seems to me that the right process is that the Defendants may still redact data and parameters of those elements it says are irrelevant, but the provisions of all the descriptions and headings should then allow focused debate, if there needs to be debate, on potential relevance in relation to those parameters which have been withheld.

The Defendants must also review those elements which it has redacted in light of the observations I have made. In light of those comments, the Defendants are to adopt a more expansive view of relevance than it has adopted to date. I say in terms now that if particular parameters are sought by the Claimants in due course, having been provided with headings and descriptions, on the basis that their technical advisers honestly believe that particular parameters are likely to be relevant to their investigations, the Court is likely to err in favour of disclosure, given that the documents are being disclosed into the confidentiality ring. The review process to be undertaken by the Defendant is also to be supervised and overseen by a solicitor, who will properly verify that the redactions made pursuant to this Order are appropriate and comply with the observations I have made. It goes without saying that the Court will not be impressed in due course by being troubled with issues of relevance if the application of my observations to date to the dispute would provide a ready answer.

There is to be no prolonged debate between the parties. If there is a dispute at all, it is to be crystallised swiftly and brought before the Court, whether at a designated progress meeting or a further specific hearing, or indeed on papers, if that is the way the parties think it can most efficiently be resolved.

In relation to the number of files to be disclosed, I consider that there is no basis to revisit the Court's two Orders. The samples (in the Mercedes GLO) had been chosen by the March CMC and no application was made then for a variation. I also accept there is likely to be relevance in testing the wider arguments around heterogeneity and, in particular, the Claimants may have to demonstrate in due course, the representative nature of the firmware, and I can see that these files may be probative in context of that question.

I also note that the other ALGLOs are being required to provide the A2L files on a wider basis and it is sensible to retain consistency. The scope of the order in this GLO shall remain undisturbed.