Locations
First appeared in Privacy & Data Protections Volume 13, Issue 3
Victoria Hordern, Senior Associate at Fieldfisher, discusses the thorny issues of consent and legitimate interests under the draft Regulation, taking into account recent amendments proposed in the LIBE Committee report.
In the context of data protection compliance, is consent the central principle? Is consent the means of giving individuals true control over their personal information? This is what the Rapporteur for the European Parliament’s Committee on Civil Liberties, Justice and HomeAffairs, Jan Philipp Albrecht, contends in his draft report on the EU Data Protection Regulation.
On the one hand, consent could be said to place individuals and their choices centre stage in data protection compliance. On the other hand, how realistic is it to really rely on consent as the ground to justify data processing particularly in the context of the online, networked world? Just how many consents would I need to give every day
when I use the internet?
The consent standard
There are a number of specific requirements that a controller has to meet in order to rely on consent as a ground to justify data processing. To pick just one, a controller needs to be confident that any consent gathered is informed — in other words, that the individual understands what they are consenting to. Regulators recognise that levels of understanding can vary, which is why they have consistently required controllers to go to greater lengths when notifying children (as opposed to adults) about what will happen to their data.
At least three practical obstacles emerge when relying on consent. Firstly, a controller is faced with the possibility that, having obtained consent, an individual argues later on that they didn't really consent. That, despite the information provided by the controller, the individual didn't receive adequate information or didn't fully understand it. Secondly, as time elapses, the question arises concerning whether the consent remains valid and whether the controller has to go back and obtain a further consent. Thirdly, the controller needs to be able to show that consent was obtained. Regarding the third point, how far evidentially does this need to go?
Problems with consent
Paradoxically, relying on consent could lead to a lower standard of data protection. If a controller obtains a broad consent (and relies on the fact that individuals will not closely scrutinise the consent wording at the time), what incentive is there on controllers to protect the data after consent has been obtained?
This is a concern raised in the context of international data transfers. In a 2006 working document (see WP114 — www.pdpjournals.com/docs/88080), the Article 29 Working Party reiterated that controllers making international data transfers should favour solutions that provide individuals with a guarantee that they will continue to benefit from the data protection rights that they are entitled to in the EU once their data have been exported outside the EU. Consequently, relying on derogations (including consent) should be a last resort, and other solutions which have a longer lasting effect (such as contracts or Binding Corporate Rules) are preferable. To adapt the words of the Working Party in WP114, it could be said that “consent is unlikely to provide an adequate long-term framework for data controllers in cases of repeated processing”.
Elsewhere, in WP187 (http://www.pdpjournals.com/docs/88081.pdf), the Working Party has analysed the different elements of consent in detail, setting out the high standard that must be reached in order to prove that valid consent has been obtained. However, even the Working Party recognised that consent is sometimes a weak basis for justifying data processing and that it loses its value when it is stretched or curtailed to make it fit situations that it was never intended for. (A closer look at the implications of the Working Party’s definition of consent is available in the article ‘Clarifying consent’, on pages 3-5 of Volume 11, Issue 8 of Privacy & Data Protection).
How widespread is consent?
Many if not most controllers currently processing data either do not rely on consent as the lawful ground or, if they do, will not reach the high standard set out by the regulators in every instance. In reality, an individual’s experience in the online world would be significantly disrupted if organisations were required to obtain their consent (as defined by the regulators) on a regular basis. Indeed, there are certain situations where it will be obvious to individuals that organisations will collect and use their personal information and consent would be superfluous.
Furthermore, where the use of personal information involves very little intrusion into an individual’s privacy, obtaining consent in such circumstances is very heavy handed.
This is not to say that consent is never the appropriate route to go down. Clearly, in situations where sensitive information is involved (for example health data), consent can be suitable. But for all practical purposes, a substantial amount of data processing that takes place currently relies on other grounds, and especially the ‘legitimate interest’ ground.
The legitimate interest ground requires a controller to demonstrate that the data processing is necessary for his legitimate interests, and that his interests are not overridden by the interests of the affected individuals. In other words, the controller is required to carry out a balancing exercise to assess whether there are any data protection risks to the individuals affected that would prevent it from pursuing its purposes.
The Rapporteur’s enthusiasm for consent
Enter Jan Philipp Albrecht, Rapporteur, who published a draft report on the EU Commission’s Data Protection Regulation on 8th January 2013. While many considered the EU Commission’s Data Protection Regulation published in January 2012 to be strict, the amendments in Albrecht’s version of the Regulation are even more so. And one area where this is clearly seen relates to the position of consent and legitimate interests. For a start, Albrecht effectively elevates consent to the primary lawful ground for data processing, something that the Working Party in its previous opinions has been careful not to do.
Where the EU Commission’s draft Regulation states that consent is not valid where an individual is in a situation of dependence upon the controller, Albrecht’s amendments extend this restriction by stating that consent is also not valid where the controller is in a dominant market position with respect to the products or services offered to the individual. Since Albrecht’s amendments also severely narrow the scope of the legitimate interest ground (see below), this restriction is problematic for online platforms that occupy key positions in their markets (cloud storage providers or social network services) and offer their services for free (so cannot rely on the contractual necessity ground to lawfully process personal information).
Legitimate interest — a last resort
According to Albrecht, the legitimate interest ground should be effectively a last resort where no other lawful grounds for data processing are available. His amendments to the legitimate interest ground aim to “provide legal certainty for data processing based on the legitimate interest of the data controller”. In pursuit of legal certainty, the amendments strictly prescribe the circumstances where the controller’s legitimate interests override the interests of the individual. The circumstances are, broadly speaking:
- journalism;
- protecting the controller’s or a third party’s legal rights
- direct marketing where there is a pre-existing contractual relationship;
- within a professional business-tobusiness relationship; and
- processing by charities.
Looking at it from the other point of view, the amendments also state that the individual’s interests override the controller’s legitimate interests in strictly prescribed circumstances:
- serious risk of damage to the individual;
- where sensitive information is used;
- where the individual can reasonably expect that his personal information will only be used for a specific purpose or treated confidentially;
- profiling;
- big data-style processing;
- where processing will lead to an adverse affect on the individual due to defamation or discrimination; and
- where the individual is a child.
Therefore, these amendments already determine in what situations a controller's interests trump an individual’s and vice versa. The balance of interests test is effectively no longer one that the controller makes — it has already been made for him. And the effect of the amendments to the legitimate interest ground do not stop there. Where a controller wishes to rely on the legitimate interest ground he must:
- inform the individuals about the processing explicitly and separately; and
- publish the reasons for believing that his interests override the interests of the individual.
The impact of these changes
The proposed amendments to the legitimate interest ground would make it even harder for non-public sector controllers to rely on any lawful ground for data processing other than contractual necessity or consent (unsurprisingly Albrecht leaves intact the provision prohibiting public authorities from relying on the legitimate interest ground).
The proposed amendments also leave little room for a flexible interpretation of legitimate interests that accounts for any actual risk to the individual’s privacy, a flexibility which is generally available under the EU Data Protection Directive.
Ultimately, the message of the Albrecht amendments is that private sector controllers should either rely on consent or contractual necessity, only look at legitimate interest if no other ground is available and, even then, only rely on the legitimate interest ground as specifically defined.
Certain industries are likely to lobby hard against these proposed changes. The prospect of having to frequently obtain consent (and meeting the specific requirements of consent laid down in WP187) will be immensely frustrating for organisations, particularly those that operate online. Data protection regulators may also object to the narrow scope proposed for relying on legitimate interests. The UK’s Information Commissioner’s Office in its February 2012 analysis of the Draft Data Protection Regulation intimated that it would not wish to see greater restrictions around the legitimate interest ground.
Conclusion
As the dust settles following the announcement of the draft report from the Rapporteur, there is a debate to be had about the realistic place of consent and legitimate interests as grounds under the new Data Protection Regulation. Is consent really the panacea to all data protection compliance problems? Or in reality, would a blizzard of consent notices lead to consent-fatigue where individuals fail to engage with what they are consenting to? How can controllers be incentivised to better protect personal information? Is predetermining the circumstances under which legitimate interests can be relied upon really the best outcome? Would giving controllers the responsibility to assess the privacy risks for themselves actually ensure better accountability for what happens to the data?
Initial noises from the EU Commission suggest that Albrecht’s proposals on consent are not what they had in mind. The next few months will be crucial in hammering out these key issues before the wording of the Regulation is finalised.
Victoria Hordern, Senior Associate in Fieldfisher's Privacy & Information Law Group