Introduction
On October 21, 2013, the European Parliament's Committee on Civil Liberties, Justice and Home Affairs (LIBE Committee) held its long-awaited vote and adopted compromise amendments that would modify the proposal of the European Commission (Commission) for a draft European Union Data Protection Regulation (Regulation).1 This alert gives background information on the vote and next steps, and provides an overview of some key proposed amendments. Although the vote is an important step towards the adoption of a new EU data protection legal framework, its significance should not be overestimated as many steps must still be completed before enactment.
Background and Next Steps
On January 16, 2013, Jan Philipp Albrecht, the rapporteur for the LIBE Committee, issued an initial report on the Commission proposal (the Albrecht Report).2 Subsequently, all members of the European Parliament were invited to submit amendments. As a result, almost 4,000 amendments were tabled, forcing the LIBE Committee to postpone its vote twice.
This vote gives the Parliament a mandate to start negotiating the text with the Council of the EU (Council), meaning the 28 EU Member States. In addition, the Council will continue working towards an initial agreement among its members.3 Once the Council has agreed internally on its own text, it will negotiate a final text with the Parliament and the Commission, which will require final approval from the Parliament sitting in plenary session. The Parliament and Commission are pushing hard to reach such a stage before the parliamentary elections in May 2014. Although the EU institutions are optimistic about finalizing the legislative process soon, adoption of a final text of the Regulation before the upcoming parliamentary elections remains uncertain.4
Analysis of Selected Compromise Amendments
While many of the compromise amendments are similar to those presented in the Albrecht Report, the text also makes some significant changes. Some of the major provisions of the compromise text include the following:
-
Extraterritorial effect. The LIBE text provides that the Regulation would apply to both controllers and processors not established in the EU when offering services to individuals in the EU, even without payment, or when monitoring such individuals ("monitoring" seems to involve tracking and the creation of profiles) (Article 3 and Recital 21). Data processors would be directly subject to the Regulation, meaning that it would apply to a variety of online service providers located outside of the EU.
-
Pseudonymous and encrypted data. The compromise text introduces new concepts with regard to the definition of personal data: (1) "pseudonymous data," defined as personal data that "cannot be attributed to a specific individual without the use of additional information," as long as such information is kept separately and secure; and (2) "encrypted data," identified as personal data that is "rendered unintelligible" to unauthorized access due to security measures (Article 4 (2a) and (2b)). The LIBE amendments clarify that such types of data remain personal data under the Regulation, but they are subject to less burdensome requirements.
-
Main establishment and one-stop shop. LIBE suggests harmonizing the concept of "main establishment" for both controllers and processors. The decisive criterion in the compromise text is the location where the main decisions are made with regard to conditions and means of the processing (Article 4(13)).5
In addition, the compromise text makes changes to the one-stop shop mechanism. The rationale for the one-stop shop approach is to have one data protection authority (DPA), such as the DPA of the company's main establishment, competent for all of its data processing activities in the EU. Contrary to the initial Parliament position that considered the lead DPA to be a mere contact point and to the initial Commission proposal that provided for a true one-stop-shop approach, the compromise text now takes an intermediary position and creates a "lead DPA" system (Article 54a).
According to the compromise text, the lead DPA would be the sole authority empowered to take legal decisions with regard to a company, but would have complex cooperation obligations with regard other DPAs. Furthermore, individuals could lodge a complaint before the DPA of their home jurisdiction, and the lead DPA would be required to coordinate its work with that DPA.
-
Data transfers and law enforcement requests. The compromise amendments provide that Commission adequacy decisions (such as the one on Safe Harbor) would remain in force for five years after the Regulation goes into effect, unless they are amended, replaced, or repealed by the Commission (Article 41(8)). There is also a sunset clause for data transfer authorizations based on Article 26(2) of the current Directive, meaning, for example, that authorizations for BCRs or Standard Contractual Clauses would have to be re-authorized by DPAs within two years of the entry into force of the Regulation (Article 42).
Furthermore, several elements that were considered to be improvements in the Commission's proposal are deleted or replaced by stricter requirements (e.g., the legitimate interest legal basis for certain data transfers has been deleted) (Article 44). However, elements such as the recognition of the European Data Protection Seal as a valid mechanism for data transfers, and the fact that any legally binding conventions or instruments dealing with data protection should be taken into account when assessing the adequacy of a third country, can be considered to be positive developments (Articles 41(2) and 42(2)).
In addition, the LIBE draft would require both data controllers and data processors to notify DPAs about requests to disclose personal data to courts or regulatory authorities in countries outside of the EU, and to obtain formal approval of DPAs before turning over European data (Article 43a). The LIBE text also provides that "any legislation which provides for extra-territorial access to personal data processed in the Union without authorization under Union or Member State law should be considered as an indication of a lack of adequacy" (Recital 82). These provisions are controversial and are likely to be debated during the further legislative process.
Other important compromise amendments include:
-
Legal basis for data processing. Additional restrictions would be imposed in order for consent to be valid. The compromise text would require companies to obtain explicit consent (pre-ticked boxes would not suffice), limit consent to specific purposes, and not make consent conditional for processing that is not necessary for the requested services (Article 7). Furthermore, the use of a company's legitimate interest as a legal basis is retained, but is further restricted. Companies could not rely on their legitimate interest to process data when it does not meet individuals' reasonable expectations (Article 6)—it is unclear what this would mean in practice, but it could certainly be used to restrict the processing of personal data. However, companies would be allowed to rely on their legitimate interest to process pseudonymous data (Recital 38) which would introduce some welcome flexibility.
-
Cookies and IP addresses. The compromise text explicitly mentions that cookies and IP addresses constitute personal data, unless such identifiers do not relate to an identified or identifiable individual (Recital 24).
-
Use of icons to provide notice. The LIBE text requires companies to complement privacy policies with icons that would describe in a graphical way a number of elements, such as how personal data is being collected, retained, and shared with third parties and how encryption is used (Article 13a).
-
Right to be forgotten. The right to be forgotten has been renamed and merged with the right to erasure. Some concerns have been taken into account but others remain, such as requesting companies that have made data public without legal justification to erase it, including data held by third parties (Recital 54). In addition, individuals can request third parties to erase any links to or copies of data or otherwise request restriction of the processing based on court order or if the particular type of storage technology no longer allows for erasure (Article 17).
-
Profiling. The LIBE compromise text restricts profiling activities when they lead to measures producing legal effects or when they significantly affect the interests, rights, and freedoms of individuals. In these situations, profiling is allowed only if it is based on individuals' consent, if provided by EU Member State law, or if conducted in the context of the performance of a contract (and if adequate safeguards are implemented). Profiling that is based solely on sensitive data is prohibited. However, some flexibility has been introduced for cases where profiling is based on pseudonymous data, provided that it is impossible for the controller to attribute the data to a specific individual based on a single source of pseudonymous data or on aggregated pseudonymous data. This could in theory introduce some flexibility for companies conducting online data analytics (Article 20 and Recital 58a).
-
Compliance and data protection officers (DPOs). The compromise text mandates a bi-annual review and update of compliance policies and procedures (Article 22), and requires companies to designate a DPO when the processing affects more than 5,000 individuals in a consecutive 12-month period (Article 35).
-
Breach notification. The 24- or 72-hour deadline to report data breaches to DPAs contained respectively in the initial Commission proposal and in the Albrecht Report has been withdrawn, and the draft now requires companies to notify "without undue delay" (Article 31).
-
Sanctions and fines. The fines have been significantly increased and can now amount to €100 million or up to 5 percent of a company's annual worldwide turnover, whichever is greater (Article 79).
Conclusions and Implications for Companies
The LIBE vote is an important step in the legislative process, but the road ahead is long and many obstacles (described above) could still slow down the possible adoption of the new EU data protection framework. In particular, it is unclear whether the Council of the EU will reach an agreement with the other institutions.
In addition, it seems that the plenary of the EU Parliament intends to vote on the data protection reform in April 2014 in order to put pressure on the Council and create a legacy with a view to resuming work after the upcoming parliamentary elections. Thus, if no agreement is reached between the EU institutions before the parliamentary elections in May 2014, work would likely resume after the new Parliament and Commission are in place in October 2014. This gives an additional indication that the legislative process is far from being completed. In any event, the Regulation would only come into effect only two years after enactment.
If the Regulation is adopted, it will significantly impact how businesses operate and process data globally. Although many details may still be significantly modified in the course of the legislative process, consensus is emerging among observers that the key principles of the Regulation will somehow be inserted in the new EU data protection framework. Therefore, companies should consider how these key principles could impact their operations, strategize on whether they will affect their data protection strategy and compliance programs, and begin planning modifications to their existing practices if necessary.
