The EU’s Data Protection Regulation: where are we?

The replacement of Directive 95/26/EC – the bedrock of data protection in Europe – with a new Regulation is intended as a radical overhaul, making protections for personal data fit for the digital world. It has now been over two years since the first substantive draft of that Regulation was made public. I dimly recall Tim Pitt-Payne and I summarising it – see here.

The Regulation is yet to emerge. As a number of Panopticon readers have asked: where have we got to? Here are five points by way of summary.

1. Two members of the trinity are on board

Following seemingly interminable negotiations, the European Parliament’s civil liberties committee (LIBE) now endorses the European Commission’s position on the modified draft. This means that two of the three key bodies at the EU level appear to be of one mind. The next step is for the third body, the European Council, to be persuaded during negotiations. See this blog post by the ICO’s Deputy Commissioner, David Smith.

2. In search of the cardinal virtues – consent, consistency, proportionality

In a very illuminating summary of the major principles at issue, the ICO tells us that it welcomes the following features of the current draft: a stringent approach to consent (or, in low-risk situations, a ‘legitimate interests’ condition justifying the processing of personal data); consistency and an EU-wide ‘one-stop shop’ model; ensuring that processing conditions are proportionate to risk (by, for example, requiring data subjects to be notified ‘without delay’ rather than within 24 hours, as was originally proposed).

The ICO remains concerned, however, that the draft Regulation continues to suffer from some vices: its use of the ‘pseudonymisation’ concept muddies the distinction between personal and non-personal data; the approach to profiling is insufficiently nuanced, and the international transfer rules may be unrealistically stringent.

3. The Regulation is dead!

Peter Fleischer, Google’s global privacy counsel, considers that the stalled progress of 2013 effectively means that “the old draft is dead”. His view, however, is that this delay will provide an opportunity for a more realistic re-think: “Whatever comes next will be the most important privacy legislation in the world, setting the global standards. I’m hopeful that this pause will give lawmakers time to write a better, more modern and more balanced law.”

4. Long live the Regulation!

EU officials are, however, optimistic about the current draft being spurred on to finality in 2014. Peter Hustinx, the outgoing European Data Protection Supervisor (curiously, no successor has yet been appointed), hopes that Greece’s imminent turn in the presidency seat will provide a fresh impetus for productive negotiation. Importantly, he sees Germany (often characterised as setting very stringent standards for data protection) as being in the driving seat: “The new German government can tackle this subject with the necessary drive and energy and thereby gain acceptance of the German position at European level and lead Europe to a higher level of data protection.”

5. Are the Americans Safe?

The processing of EU citizens’ data by US-based companies sits outside the direct reach of the envisaged Regulation, as with the current Directive. Since 2000, transfers of personal data to the US have been governed by the Safe Harbour Agreement, under which approximately 3,300 companies have been certified as safe (in the sense of being EU compliant in their data protection standards).

The European Council and Parliament have, however, expressed concern about the fitness for purpose of the Safe Harbour scheme. They have observed that “Web companies such as Google, Facebook, Microsoft, Apple, Yahoo have hundreds of millions of clients in Europe and transfer personal data for processing to the US on a scale inconceivable in the year 2000 when the Safe Harbour was created”. They area also concerned about the ongoing revelations about surveillance: “divergent responses of data protection authorities to the surveillance revelations demonstrate the real risk of the fragmentation of the Safe Harbour scheme and raise questions as to the extent to which it is enforced”.

Progress by the US Department of Commerce is now sought – by March 2014 – on improving transparency, the application of EU principles and enforcement. The arrangements will be further reviewed in 2014.

Robin Hopkins @hopkinsrobin

George Osborne, Nigel Lawson and FOIA – political vs official information

Government ministers wear two hats (apart from Vince Cable – he seems to like hats, and probably has quite a few). They are public officials, but they are also party politicians. Both of those activities are likely to generate recorded information. FOIA extends to the official information, but not the party political. This is well established in principle, but not straightforward to apply, since the two categories will often overlap. It is also surprisingly untested before Tribunals. Michael Gove was due to test the principle in a 2012 appeal, but that was withdrawn.

The issue has now been considered by the Tribunal in Brendan Montague v IC and HM Treasury (EA/2013/0074): 029 070114 Final Decision EA-2013-0074. The information in dispute was a record of a telephone conversation which took place on a Sunday morning in September 2011 between the Chancellor of the Exchequer, George Osborne, and one of his predecessors, Lord Lawson.

The ICO’s position (and that of HMT) was that some of that information was predominantly party political in nature and was thus not held by HM Treasury for FOIA purposes. The remainder was exempt under section 35(1)(a), i.e. insofar as official business was being discussed, it related to the formulation or development of government policy, and the public interest favoured maintaining the exemption.

The Tribunal disagreed on the first point: while it accepted the principle, it had “no hesitation” in concluding that all of the disputed information in this case was held by HMT for official purposes rather than Mr Osborne’s party-political ones. It was not attracted by dissecting and partitioning the record between the party-political and the official in this instance, and it favoured a restrictive approach to a principle by which information could be taken outside of FOIA’s reach.

On the section 35(1)(a) point, the Tribunal agreed that there was a need for a safe space, given the high-level economic policy issues – including concerning the banking sector – which were being discussed. It was satisfied that the disputed information did not indicate that any impropriety or lobbying was at play.

I appeared for the ICO; my colleague Julian Milford appeared for HMT. No further analysis from me, given my involvement in the case, but I post it here because of the relative novelty of the political/official information point which, one suspects, will rear its head in other cases in future.

Robin Hopkins @hopkinsrobin

Personal data and fitness to practice investigations – Tribunal overturns ‘neither confirm nor deny’ position

When an identifiable individual has been the subject of a formal complaint about their competence or conduct, that fact constitutes their personal data. In terms of privacy/publicity decisions, such situations are often approached in this way: where the complaint is well founded or at least merits serious consideration, publication is warranted, but otherwise confidentiality is maintained, lest unjustified aspersions be cast against that person.

In that respect, the process outlined by the Tribunal in Foster v IC (EA/2013/0176) – which concerned a complaint to the Nursing & Midwifery Council – is typical:

“The complaints procedure administered by the NMC has two stages. The first stage is designed to determine whether or not the matter should be referred to the NMC’s Fitness to Practice Panel. If it is, then the Panel will meet in public and its decision will be made publicly available. But if the complaint does not proceed beyond the first stage, (either because a decision is made not to investigate or because the NMC’s Investigating Committee Panel concludes that the complaint does not justify a reference to the Fitness to Practice Panel), then the process remains confidential. The rationale appears to be that an individual’s professional reputation should not be undermined by the publication of allegations that are found not to have sufficient merit to justify being referred to the Fitness to Practice Panel”.

The Appellant, whose son died following his participation in a drug trial, considered that the NMC investigation in this case – which did not pass the first stage – may have been inadequate. She asked for information about its investigation into her complaint about a named practitioner.

The NMC adopted a ‘neither confirm nor deny’ position under section 40(5), i.e. it considered that to say whether or not it held information on a complaint about this individual would be to tell the world at large whether or not that person had been the subject of a professional complaint of this description. The ICO agreed, but the Tribunal overturned that decision, ordering the NMC to confirm or deny whether it held the requested information.

In reaching that view, the Tribunal – while not passing judgment on the merits of the complaint or the NMC’s investigation – considered the criticisms that had been made:

“If it were to be the case that any member of the care team had realised the error earlier, but had not raised the alarm until after its very sad consequences had become clear, then there would seem to us to be strength in the Appellant’s argument that the evidential basis for the decision of the NMC’s Investigating Committee Panel required investigation”.

In those circumstances, the Tribunal thought the fairness balance favoured confirming or denying whether the requested information was held:

“In reaching that conclusion we reject the Information Commissioner’s argument that it is always unfair, and therefore in breach of the Data Protection Principles, to make a statement that discloses the existence of a complaint of professional misconduct against an individual, where there has been no finding of wrongdoing or malpractice. That would create an inflexible test which prevented all relevant circumstances being taken into account. Nor do we accept the Information Commissioner’s argument that the limited degree of disclosure involved in a “confirm or deny” response would constitute unwarranted interference into X’s privacy, without satisfying a legitimate public interest in disclosure”.

Public authorities who routinely adopt a default ‘neither confirm nor deny stance’ of the type outlined at the start of this post will wish to note that, at least in some circumstances, that approach can be called into question.

Robin Hopkins @hopkinsrobin

The Google/Safari users case: a potential revolution in DPA litigation?

I posted earlier on Tugendhat J’s judgment this morning in Vidal-Hall and Others v Google Inc [2014] EWHC 13 (QB). The judgment is now available here – thanks as ever to Bailii.

This is what the case is about: a group of claimants say that, by tracking and collating information relating to their internet usage on the Apple Safari browser without their consent, Google (a) misused their private information (b) breached their confidences, and (c) breached its duties under the Data Protection Act 1998 – in particular, under the first, second, sixth and seventh data protection principles. They sought damages and injunctive relief.

As regards damages, “what they claim damages for is the damage they suffered by reason of the fact that the information collected from their devices was used to generate advertisements which were displayed on their screens. These were targeted to their apparent interests (as deduced from the information collected from the devices they used). The advertisements that they saw disclosed information about themselves. This was, or might have been, disclosed also to other persons who either had viewed, or might have viewed, these same advertisements on the screen of each Claimant’s device” (paragraph 24).

It is important to note that “what each of the Claimants claims in the present case is that they have suffered acute distress and anxiety. None of them claims any financial or special damage. And none of them claims that any third party, who may have had sight of the screen of a device used by them, in fact thereby discovered information about that Claimant which was detrimental” (paragraph 25).

The Claimants needed permission to serve proceedings on the US-based Google. They got permission and served their claim forms. Google then sought to have that service nullified, by seeking an order declaring that the English court has no jurisdiction to try these particular claims (i.e. it was not saying that it could never be sued in the English courts).

Tugendhat J disagreed – as things stand, the claims will now progress before the High Court (although Google says it intends to appeal).

Today’s judgment focused in part on construction of the CPR rules about service outside of this jurisdiction. I wanted to highlight some of the other points.

One of the issues was whether the breach of confidence and misuse of private information claims were “torts”. Tugendhat J said this of the approach: “Judges commonly adopt one or both of two approaches to resolving issues as to the meaning of a legal term, in this case the word “tort”. One approach is to look back to the history or evolution of the disputed term. The other is to look forward to the legislative purpose of the rule in which the disputed word appears”. Having looked to the history, he observed that “history does not determine identity. The fact that dogs evolved from wolves does not mean that dogs are wolves”.

The outcome (paragraphs 68-71): misuse of private information is a tort (and the oft-cited proposition that “the tort of invasion of privacy is unknown in English law” needs revisiting) but breach of confidence is not (given Kitetechnology BV v Unicor GmbH Plastmaschinen [1995] FSR 765).

Google also objected to the DPA claims being heard. This was partly because they were raised late; this objection was dismissed.

Google also said that, based on Johnson v MDU [2007] EWCA Civ 262; (2007) 96 BMLR 99, financial loss was required before damages under section 13 of the DPA could be awarded. Here, the Claimants alleged no financial loss. The Claimants argued against the Johnson proposition: they relied on Copland v UK 62617/00 [2007] ECHR 253, argued for a construction of the DPA that accords with Directive 95/46/EC as regards relief, and argued that – unlike in Johnson – this was a case in which their Article 8 ECHR rights were engaged. Tugendhat J has allowed this to proceed to trial, where it will be determined: “This is a controversial question of law in a developing area, and it is desirable that the facts should be found”.

If the Johnson approach is overturned – i.e. if the requirement for financial loss is dispensed with, at least for some types of DPA claim – then this could revolutionise data protection litigation in the UK. Claims under section 13 could be brought without claimants having suffered financially due to the alleged DPA breaches they have suffered.

Tugendhat went on to find that there were sufficiently serious issues to be tried here so as to justify service out of the jurisdiction – it could not be said that they were “not worth the candle”.

Further, there was an arguable case that the underlying information was, contrary to Google’s case, “private” and that it constituted “personal data” for DPA purposes (Google say the ‘identification’ limb of that definition is not met here).

Tugendhat was also satisfied that this jurisdiction was “clearly the appropriate one” (paragraph 134). He accepted the argument of Hugh Tomlinson QC (for the Claimants) that “in the world in which Google Inc operates, the location of documents is likely to be insignificant, since they are likely to be in electronic form, accessible from anywhere in the world”.

Subject to an appeal from Google, the claims will proceed in the UK. Allegations about Google’s conduct in other countries are unlikely to feature. Tugendhat J indicated a focus on what Google has done in the UK, to these individuals: “I think it very unlikely that a court would permit the Claimants in this case to adduce evidence of what Mr Tench refers to as alleged wrongdoing by Google Inc against other individuals, in particular given that it occurred in other parts of the world, governed by laws other than the law of England” (paragraph 47).

Robin Hopkins @hopkinsrobin

High Court to hear Safari users’ privacy claim against Google

Panopticon has from time to reported on Google’s jurisdictional argument when faced with privacy/data protection actions in European countries: it tends to argue that such claims should be dismissed and must be brought in California instead. This argument is not always successful.

The same jurisdictional argument was advanced before Mr Justice Tugendhat in response to a claim brought by a group calling itself ‘Safari Users Against Google’s Secret Tracking’ who, as their name suggests, complain that Google unlawfully gathers data from Safari browser usage.

This morning, Mr Justice Tugendhat dismissed that jurisdictional argument. The case can be heard in the UK. Matthew Sparkes reports in the Daily Telegraph that the judge said “I am satisfied that there is a serious issue to be tried in each of the claimant’s claims for misuse of private information” and that “the claimants have clearly established that this jurisdiction is the appropriate one in which to try each of the above claims”.

The same article says that Google will appeal. This follows Google’s announcement yesterday that it will appeal a substantial fine issued by the French data protection authority for unlawful processing (gathering and storing) of user data.

Panopticon will continue to gather data on these and other Google-related matters.

Robin Hopkins @hopkinsrobin

UCAS and the extent of FOIA: Tribunal favours wide approach

Transparency advocates often express frustration at the number of bodies which are not within the scope of FOIA, because they are not listed or designated as ‘public authorities’ for FOIA purposes. The Coalition government responded by announcing, in January 2011, that FOIA would be extended to a number of additional bodies. This was done with effect from 1 November 2011, through the Freedom of Information (Designation as Public Authorities) Order 2011. This brought the Association of Chief Police Officers of England, Wales and Northern Ireland (ACPO); the Financial Ombudsman Service and the Universities and Colleges Admissions Service (UCAS) within the scope of FOIA.

As regards UCAS, the difficulty is that this was not done in a straightforward blanket way. In recognition of the diversity of UCAS’ functions, its amenability to FOIA was limited to information relating to the “provision and maintenance of a central applications and admissions service”. This frames UCAS’ duties in a positive way.

This is similar – but not the same as – the approach taken to the BBC, which is subject to FOIA “in respect of information held for purposes other than those of journalism, art or literature”. This frames the BBC’s duties in a negative way.

The Supreme Court in BBC v Sugar (No2) told us how to approach the extent of the BBC’s FOIA duties. How should Sugar be applied to the differently-worded UCAS provision?

This was the issue before the Tribunal in University and College Admission Service v IC and Lord Lucas (EA/2013/0124), the requester (the author of the Good Schools Guide) made a number of requests to UCAS about university admissions. Some were refused on section 12 (cost of compliance) grounds; the ICO agreed with UCAS that the remaining information was exempt under section 43(2) (prejudice to commercial interests). UCAS and the ICO disagreed, however, about the extent to which UCAS was subject to FOIA.

UCAS argued that Sugar required the Tribunal to consider whether the information was held, to any significant degree, for a purpose other than the designation (in particular, UCAS’s commercial functions), and if so, it fell outside the scope of FOIA.

The ICO argued that because the BBC and UCAS were in reverse positions (the BBC being subject to a specific exclusion, and UCAS subject to a specific inclusion), the question should be whether the information was held to any significant degree for the designated purpose, and if so, it fell within the scope of FOIA. Both parties argued that the other was turning Sugar on its head.

The Tribunal adopted the ICO’s analysis of Sugar. The primary purpose of the 2011 Order was to bring UCAS within the scope of FOIA and subject it to the principles of greater openness and transparency that such a designation was designed to bring: at [68]. The focus of the phrase “the provision and maintenance of a central applications and admissions service”, taken with section 7(5) FOIA, is on what is actually caught by FOIA and the purpose of that wording is specifically to include information: at [66].

In favouring this wider approach to the application of FOIA to UCAS, the Tribunal said this:

“71. Most persuasive is the IC’s point that, in construing the scope of the 2011 Designation Order, it is important to recall that Parliament would have been well aware of the existing exemptions provided in FOIA. There is no need to read the 2011 Designation Order narrowly to ensure there is no overlap with a commercial function of UCAS because section 43 FOIA itself provides protection to UCAS in relation to information which prejudices its commercial interests.

72. The approach of UCAS in this case would have the result that only admissions data relating to the currently live admissions round would fall within the scope of FOIA. This surprisingly narrow result is unlikely to have been the one intended by Parliament when designating UCAS as a public authority for FOIA, not least because the ‘”provision and maintenance of a central applications and admissions service” does not suggest such an outcome.”

11KBW’s Chris Knight appeared for the ICO.

Robin Hopkins @hopkinsrobin