“Humans were always far better at inventing tools than using them wisely.” The President of the European Commission used this quote from Yuval Noah Harari, the author of “Sapiens”, when she presented her new package of measures for the technologies of the future on Wednesday 19 February.
A core part of the proposals are new provisions designed to facilitate data sharing, which is the basis of genuine artificial intelligence, a stricter framework for high-risk artificial intelligence, a reflection on facial recognition and a series of announcements concerning European digital policy. All of these ideas are contained in three separate documents - a general strategy (http://bit.ly/2v1IAwm ), a data strategy (http://bit.ly/38HHOn4 ) and a consultation on artificial intelligence (http://bit.ly/2v1Igh8 ) - accompanied by two expert reports: one on the safety and liability implications of artificial intelligence, the Internet of Things and robotics (http://bit.ly/2ubmzuI ) and the other on data sharing between business and government (http://bit.ly/39Lbq2N ).
Key changes
The most important change since our previous report (see EUROPE 12427/17) is the European Commission's commitment to find €2 billion (instead of €1 billion) to fund data spaces and federated cloud infrastructures. EUROPE has reviewed the Commission's main commitments and the changes to previous leaked versions, which can be summarised as follows:
- Artificial Intelligence. The Commission has retained its risk-based approach (see EUROPE 12415/1). It believes that conformity assessments will be necessary for applications in certain key sectors where use of AI poses a risk to human rights. “The prior conformity assessment could include procedures for testing, inspection or certification. It could include checks of the algorithms and of the data sets used in the development phase”, says the document, which suggests a voluntary labelling scheme for other applications. The Commission also suggests establishing a European governance structure to facilitate cooperation between national authorities on this issue.
On the sensitive issue of facial recognition, Margrethe Vestager, the Vice-President of the European Commission, said at a press conference: “In some instances, facial recognition will be harmless. It may also be used for remote identification, in public spaces. In this case, it will not necessarily be your face that is examined; it could be the way you walk. It’s a different cup of tea”, she explained, stressing that the document is intended to start a dialogue on the specific circumstances, if any exist, that might justify the use of this technology in public spaces.
The public consultation will be open until 19 March 2020, and the Commission aims to submit legislative proposals by the end of the year.
- Data strategy. The Commission is convinced that the potential for the data created in Europe remains largely untapped, so it suggests making it easier to store and share data, as data is essential for training artificial intelligence systems.
Compared with the draft text we disclosed at the beginning of February (see EUROPE 12418/1), the Commission is reviewing the amount of funding it is prepared to release. It is expecting public and private co-investment (Member States, the Commission and industry) of up to €4-6 billion, of which the Commission could contribute up to €2 billion in proportion to what is agreed in budget negotiations.
The Commission has retained its idea of creating (metaphorical) sectoral data spaces and creating a governance framework in the fourth quarter of 2020 (as opposed to the third quarter).
Future reforms
In the general strategy, the European Commission states that it is at a fork in the road with regard to competition rules and the responsibility of platforms.
- Reviewing competition rules. With regard to the first matter, the Commission notes that it is currently conducting an evaluation and review of the fitness of EU competition rules for the digital age (until 2023), particularly the rules governing horizontal and vertical agreements and of the market definition notice, and a “fitness” check of various state aid guidelines. It also announced its plan to launch “a sector inquiry with a strong focus on these new and emerging markets that are shaping our economy and society”. When she was asked about this at the press conference, Vestager said that the sectors in question, “where the economy is developing and where there are new issues”, had not yet been decided. The Commission has therefore committed itself to explore, in the context of the Digital Services Act package, “ex ante rules to ensure that markets characterised by large platforms with significant network effects acting as gatekeepers, remain fair and contestable for innovators, businesses, and new market entrants”. Speaking in English, Thierry Breton, the Commissioner for the Internal Market, was more assertive: “Some platforms are gatekeepers and that's not what we want for our internal market - here we will have regulation ex ante and we will inform them”, he said.
- Responsibility of platforms. With regard to the second matter, online platforms, we know that the Commission intends to submit legislation on digital services. However, the details of the initiative still need to be set out, said Vestager, indicating that she is fully committed to the new technology package. “I will always prefer soft law and shared governance to regulation. But if the platforms don't find the way to behave that we think we need, we will have to regulate in the Digital Services Act”, said Thierry Breton. (Original version in French by Sophie Petitjean)