My firm Mishcon de Reya is acting for John Pring, stalwart editor of Disability News Service, who has been seeking access to his personal data from DWP for more than a year. The ICO upheld his complaint but (see this blog, passim) said it wouldn’t take steps to require DWP to comply.
As a result of the latest letter, and media coverage, ICO has said it is reopening the case.
Filed under access to information, DWP, GDPR, human rights, Information Commissioner, subject access, UK GDPR
A while ago I wrote a piece on the Mishcon de Reya website pointing out that the Information Commissioner’s Office (ICO) had silently changed its guidance on how to calculate the “one month” timescale for responding to a subject access request under the General Data Protection Regulation (or “GDPR” – which is now domestic law in the form of the amended retained version of the GDPR, aka “UK GDPR”).
The nub of that piece was that the ICO (following the legal precedents) was now saying that “You should calculate the time limit from the day you receive the request“. Which was a change from the previous position that “You should calculate the time limit from the day after you receive the request “.
I have noticed, however, that, although the ICO website, in its UK GDPR guidance, maintains that the clock starts from the date of receipt, the guidance on “Law Enforcement Processing” (which relates to processing of personal data by competent authorities for law enforcement purposes under part 3 of the Data Protection Act 2018 (DPA), which implemented the Law Enforcement Directive) states that the time should be calculated
from the first day after the request was received
It’s not inconceivable (in fact I am given to understand it is relatively common) that a some controllers might receive a subject access request (or other data subject request) which must be dealt with under both the UK GDPR and the Law Enforcement Processing provisions (police forces are a good example of this). The ICO’s position means that the controller must calculate the response time as starting, on the one hand, on the date of receipt, and, on the other hand, on the day after the date of receipt.
And if all of this sounds a bit silly, and inconsequential, I would argue that it is certainly the former, but not necessarily the latter: failure to comply within a statutory timescale is a breach of a statutory duty, and therefore actionable, at least in principle. If the ICO really does believe that the timescale works differently under different legal schemes, then how, for instance can it properly determine (as it must, when required to) under Articles 57(1)(f) and 77(1) of the UK GDPR, or section 51(2) of the DPA, whether there has been a statutory infringement?
Statutory infringements are, after all, potentially actionable (in this instance either with regulatory action or private action by data subjects) – the ICO maintains a database of complaint cases and publishes some of this (albeit almost two years in arrears), and also uses (or may use) it to identify trends. If ICO finds that a controller has made a statutory infringement, that is a finding of potential significance: if that same finding is based on an unclear, and internally contradictory, interpretation of a key aspect of the law, then it is unlikely to be fair, and unlikely to be lawful.
The views in this post (and indeed most posts on this blog) are my personal ones, and do not represent the views of any organisation I am involved with.
UPDATE: 23.08.20 In this piece Chris Pounder identifies what the government sees as a justification for the exam scripts exemption. In a document prepared to assist adequacy discussions with the European Commission, it is said that the exemption “aims to protect the integrity of exams by ensuring that exam scripts cannot be accessed outside established processes” (on the basis that exam boards often re-use or re-purpose exam questions). However, and as Chris implies, this simply isn’t sufficient to justify the blanket exemption, not the breadth of its scope. Moreover the ICO’s meek acceptance that it permits an interpretation which even covers assignments and, presumably, other coursework, is deeply disappointing. END UPDATE.
Domestic data protection law says that students can’t later access data recorded by themselves during an exam or assessment. Why is that? And is it compatible with the UK’s obligations under GDPR and more general human rights law?
As is well known, the General Data Protection Regulation (GDPR) has direct effect on member states of the EU. This is, however, subject to certain provisions which allow member states to legislate for specific exemptions or restrictions. An example is Article 23 of GDPR, which allows member states to restrict by way of a legislative measure the scope of certain data subject rights, including the right of access at Article 15. Such restrictions must, though, respect “the essence of the fundamental rights and freedoms” and be a “necessary and proportionate measure in a democratic society” to safeguard, among a list of things, important objectives of general public interest.
The specific UK restrictions made in respect of Article 23 lie primarily in Schedule 2 of the Data Protection Act 2018. Of particular interest at the current time is the Schedule 2, paragraph 25(1) exemption to the Article 15 right of subject access which says that the right does “not apply to personal data consisting of information recorded by candidates during an exam” (and paragraph 25(4) says that “‘exam’ means an academic, professional or other examination used for determining the knowledge, intelligence, skill or ability of a candidate and may include an exam consisting of an assessment of the candidate’s performance while undertaking work or any other activity”).
Thus it is that guidance from the Information Commissioner’s Office (ICO) says, in relation to this year’s exam awards
The exam script exemption applies to information that has been recorded by the students themselves during an exam or assessment. Therefore students do not have a right to get copies of their answers from mock exams or assignments used to assess their performance
But why does this exemption exist? Search me. Why did it also exist in the 1998 Data Protection Act? Also, search me. Also search Hansard, like I have done, and you may struggle to find out. (Please let me know if I’ve missed something).
So in what way can the exam script exemption be said to respect the essence of the fundamental rights and freedoms and be a necessary and proportionate measure in a democratic society? Is this a case where Parliament merely nodded through a provision which it also merely nodded through 22 years ago?
Note that this is not a question as to whether information recorded by candidates during an exam is their personal data. It most certainly is, as the CJEU found in 2017 in Nowak. But note also that the court, in that case, observed that “the use of [such] information, one consequence of [the use of the information] being the candidate’s success or failure at the examination concerned, is liable to have an effect on his or her rights and interests, in that it may determine or influence, for example, the chance of entering the profession aspired to or of obtaining the post sought”. The court also noted, in holding that such information was personal data, the importance of the data subject’s rights of access, rectification and objection.
And let us remember recital 63 GDPR, which reminds us that one purpose of the right of subject access is to be able to “verify the lawfulness of the processing”. In the absence of any indication as to why the UK decided to restrict the right of access in such a way as to prevent students, especially this year’s students, accessing their own assignment and mock exam data, one must query how those students can adequately verify the lawfulness of the processing by those who determined their grades.
P.S. there is an argument that the ICO should do something about this, under its Article 57 tasks to monitor and enforce GDPR, to handle complaints from data subjects, and to advise parliament, the government, and other institutions and bodies. It has the power under Article 58 to issue an opinion to those bodies.
The views in this post (and indeed most posts on this blog) are my personal ones, and do not represent the views of any organisation I am involved with.
(See also the follow-up piece here)
An interesting case is being heard in the High Court, arising from an apparent error whereby, in responding to a subject access request (SAR), the London Borough of Lambeth allowed the recipient (and now defendant) data subject to electronically manipulate the information sent to him. This in turn enabled him to remove redactions, and identify someone who had made allegations against him and his wife (about the care they were providing to their child).
This is nightmare scenario for a controller – to inadvertently disclose extremely sensitive information, while responding to a SAR. In this instance, Lambeth have now brought a claim in breach of confidence against the defendant data subject, on the grounds that: the data was provided to the data subject in circumstances where he knew it was confidential; that he breached that confidentiality by unredacting the data, retaining an unredacted copy of the file, using the evidence to write a pre-action letter to the person who made allegations against him and his wife and threatening to bring court proceedings against them based on the information; and that it is integral to the work of Children’s Services that people who bring to its attention instances of perceived inadequate care or neglect of children are able to do so under conditions of confidentiality and can be assured that their confidentiality will be respected.
The instant proceedings were primarily concerned with a strike-out application by the defendant data subject, on the grounds of non-compliance by Lambeth with its (litigation) disclosure obligations. This application was roundly dismissed, and the matter will proceed to trial.
But of particular note is that, notwithstanding that the original error was Lambeth’s, it was revealed in the proceedings that the Information Commissioner’s Office (ICO) is also prosecuting the defendant data subject on charges of committing the offences of knowingly or recklessly re-identifying de-identified personal data, without the consent of the data controller, and knowingly or recklessly processing re-identified personal data, without the consent of the data controller. These are new offences created by sections 171(1) and 171(5) of the Data Protection Act 2018, and, when that Act was passed, it appeared that the mischief the provisions sought to address was the risk of hackers and fraudsters attempting to identify data subjects from large datasets (see the debates at Bill stage). It will be interesting to see if the ICO’s prosecution here results in a conviction. But it will also be interesting to see if ICO considers similar prosecutions in other circumstances. Although there is a public interest defence (among others) to section 171 charges, it is not an uncommon occurrence for public authorities (particularly) to inadvertently disclose or publish information with imperfect redactions. It certainly appears, on a plain reading of section 171, that someone re-identifying de-identified personal data (even if, say, for idle reasons of curiosity) might not always be able to avail themselves of the public interest defence.
And what is unsaid in the judgment, is whether Lambeth are facing any sort of civil, regulatory action from the ICO, arising from their error in sending the imperfectly redacted information in the first place.
The views in this post (and indeed most posts on this blog) are my personal ones, and do not represent the views of any organisation I am involved with.
A thread on Twitter by solicitor Martin Sloan has drawn attention to a change to official guidance on the question of when a subject access request (pursuant to Article 15 of the General Data Protection Regulation (GDPR)) “starts”, in circumstances where a controller processes large amounts of data and asks the data subject to specify what information is sought.
Recital 63 of GDPR says that where a controller processes “a large quantity of information concerning the data subject [it] should be able to request that, before the information is delivered, the data subject specify the information or the processing activities to which the request relates”. This certainly seems to suggest that it is only when the controller is ready to “deliver” the information (i.e. when it has already searched for and retrieved it) that it can ask for the request to be, in effect, narrowed down.
However, guidance from the Information Commissioner’s Office (ICO) used to say* “If you process a large amount of information about an individual you can ask them for more information to clarify their request. You should only ask for information that you reasonably need to find the personal data covered by the request. You need to let the individual know as soon as possible that you need more information from them before responding to their request. The period for responding to the request begins when you receive the additional information” (emphasis added). This was similar to the position which obtained under the prior Data Protection Act 1998, which provided that a controller was not obliged to comply with a request unless it was supplied with such information as was reasonably required to locate the information which the data subject sought.
But the ICO now says: “If you process a large amount of information about an individual, you may ask them to specify the information or processing activities their request relates to before responding to the request. However, this does not affect the timescale for responding – you must still respond to their request within one month” (emphasis also added).
The change appears to be correct as a matter of law (by reference to recital 63), but it is possible that it may lead to an increase in reliance by controllers on Article 12(3), which potentially allows an extension to the one month period for compliance if a request is complex.
The new wording is contained in the ICO’s draft detailed guidance on subject access requests, which is currently out for consultation. One presumes the ICO thought this particular change was sufficiently important to introduce it in advance, but it is rather surprising that no announcement was made.
[UPDATE: Martin has now got a piece on Brodies’ own website about this].
[*the link here is to an archived page].
Filed under Data Protection, GDPR, Information Commissioner, subject access
Pre-ticked consent boxes and unsolicited emails from the Consumers’ Association
Which?, the brand name of the Consumers’ Association, publishes a monthly magazine. In an era of social media, and online reviews, its mix of consumer news and product ratings might seem rather old-fashioned, but it is still (according to its own figures1) Britain’s best-selling monthly magazine. Its rigidly paywalled website means that one must generally subscribe to get at the magazine’s contents. That’s fair enough (although after my grandmother died several years ago, we found piles of unread, unopened even, copies of Which? She had apparently signed up to a regular Direct Debit payment, probably to receive a “free gift”, and had never cancelled it: so one might draw one’s own conclusion about how many of Which?’s readers are regular subscribers for similar reasons).
In line with its general “locked-down” approach, Which?’s recent report into the sale of personal data was, except for snippets, not easy to access, but it got a fair bit of media coverage. Intrigued, I bit: I subscribed to the magazine. This post is not about the report, however, although the contents of the report drive the irony of what happened next.
As I went through the online sign-up process, I arrived at that familiar type of page where the subject of future marketing is broached. Which? had headlined their report “How your data could end up in the hands of scammers” so it struck me as amusing, but also irritating, that the marketing options section of the sign-in process came with a pre-ticked box:
As guidance from the Information Commissioner’s Office makes clear, pre-ticked boxes are not a good way to get consent from someone to future marketing:
Some organisations provide pre-ticked opt-in boxes, and rely on the user to untick it if they don’t want to consent. In effect, this is more like an opt-out box, as it assumes consent unless the user clicks the box. A pre-ticked box will not automatically be enough to demonstrate consent, as it will be harder to show that the presence of the tick represents a positive, informed choice by the user.
The Article 29 Working Party goes further, saying in its opinion on unsolicited communications for marketing purposes that inferring consent to marketing from the use of pre-ticked boxes is not compatible with the data protection directive. By extension, therefore, any marketing subsequently sent on the basis of a pre-ticked box will be a contravention of the data protection directive (and, in the UK, the Data Protection Act 1998) and the ePrivacy directive (in the UK, the Privacy and Electronic Communications (EC Directive) Regulations 2003 (PECR)).
Nothwithstanding this, I certainly did not want to consent to receive subsequent marketing, so, as well as making a smart-arse tweet, I unticked the box. However, to my consternation, if not my huge surprise, I have subsequently received several marketing emails from Which? They do not have my consent to send these, so they are manifestly in contravention of regulation 22 of PECR.
It’s not clear how this has happened. Could it be a deliberate tactic by Which? to ignore subscribers’ wishes? One presumes not: Which? says it “exists to make individuals as powerful as the organisations they deal with in their daily live” – deliberately ignoring clear expressions regarding consent would hardly sit well with that mission statement. So is it a general website glitch – which means that those expressions are lost in the sign-up process? If so, how many individuals are affected? Or is it just a one-off glitch, affecting only me?
Let’s hope it’s the last. Because the ignoring or overriding of expressions of consent, and the use of pre-ticked boxes for gathering consent, are some of the key things which fuel trade in and disrespect for personal data. The fact that I’ve experience this issue with a charity which exists to represent consumers, as a result of my wish to read their report into misuse of personal data, is shoddy, to say the least.
I approached Which? for a comment, and a spokesman said:
We have noted all of your comments relating to new Which? members signing up, including correspondence received after sign-up, and we are considering these in relation to our process.
I appreciate the response, although I’m not sure it really addresses my concerns.
1Which? Annual Report 2015/2016
The views in this post (and indeed all posts on this blog) are my personal ones, and do not represent the views of any organisation I am involved with.
Filed under consent, Data Protection, Directive 95/46/EC, Information Commissioner, marketing, PECR, spam, subject access
Government page on subject access rights is not accurate
Right of access to data about oneself is recognised as a fundamental right (article 8(2) of the Charter of Fundamental Rights of the European Union*). Section 7 of the UK’s Data Protection Act 1998 (DPA) gives expression to this, and provides that as a general right individuals are entitled to be told whether someone else is processing their data, and why, and furthermore (in terms) to be given a copy of that data. The European General Data Protection Regulation retains and bolsters this right, and recognises its importance by placing it in the category of provisions non-compliance with which could result in an administrative fine for a data controller of up to €20m or 4% of turnover (whichever is higher).
So subject access is important, and this is reflected in the fact that it is almost certainly the most litigated of provisions of the DPA (a surprisingly under-litigated piece of legislation). Many data controllers need to commit significant resources to comply with it, and the Information Commissioner’s Office (ICO) produced a statutory code of practice on the subject in 2014.
But it is not an absolute right. The DPA explains that there are exemptions to the right where, for instance, compliance would be likely to prejudice the course of criminal justice, or national security, or, in the case of health and social care records, would be likely to cause serious harm to the data subject or another person. Additionally the DPA recognises that, where complying with a subject access request would involve disclosing information about another individual, the data controller should not comply unless that other person consents, or unless it “is reasonable in all the circumstances to comply with the request without the consent of the other individual” (section 7(4) DPA).
But this important caveat (the engagement of the parallel rights of third parties) to the right of subject access is something which is almost entirely omitted in the government’s own web guidance regarding access to CCTV footage of oneself. It says
The CCTV owner must provide you with a copy of the footage that you can be seen in. They can edit the footage to protect the identities of other people.
The latter sentence is true, and especially in the case where footage captures third parties it is often appropriate to take measures to mask their identities. But the first sentence is simply not true. And I think it is concerning that “the best place to find government services and information” (as gov.uk describes itself) is wrong in its description of a fundamental right.
A data controller (let’s ignore the point that a “CCTV owner” might not necessarily be the data controller) does not have an unqualified obligation to provide information in response to a subject access request. As anyone working in data protection knows, the obligation is qualified by a number of exemptions. The page does allude to one of these (at section 29 of the DPA):
They can refuse your request if sharing the footage will put a criminal investigation at risk
But there are others – and the ICO has an excellent resource explaining them.
What I don’t understand is why the gov.uk page fails to provide better (accurate) information, and why it doesn’t provide a link to the ICO site. I appreciate that the terms and condition of gov.uk make clear that there is no guarantee that information is accurate, but I think there’s a risk here that data subjects could gather unreasonable expectations of their rights, and that this could lead to unnecessary grievances or disputes with data controllers.
Gov.uk invite comments about content, and I will be taking up this invitation. I hope they will amend.
The views in this post (and indeed all posts on this blog) are my personal ones, and do not represent the views of any organisation I am involved with.
Filed under Data Protection, Information Commissioner, subject access
informationrightsandwrongs 