Internal Market Commissioner Thierry Breton told MEPs in the hemicycle of the European Parliament on Monday 10 February that the aim is to create a climate of confidence in Artificial Intelligence without over-regulating it. His speech followed a debate requested by MEPs (written question) before they adopted their position on automated decision-making processes.
“I don't know many Parliaments in the world where you can have a debate of this quality”, said Thierry Breton. “Beyond our partisan differences, we all shared the same questions, the same wishes. In this evolution, what we share is the place of humanity, the respect of our values, the integrity of our decisions and the understanding of the phenomenon that surrounds us”, he said to close the discussion.
The Commission called upon before the presentation of its package on February 19
The exchange of views originated in a written question drafted by the Committee on the Internal Market (IMCO) which called on the Commission to clarify how it intends to protect European consumers, in particular from unfair or discriminatory commercial practices, lack of transparency in the decision-making process and bias in data sets.
In his reply, Thierry Breton recalled that the Commission is currently preparing its Digital Europe package, namely a Communication on future challenges (see EUROPE 12419/11), a Communication on data (see EUROPE 12418/1) and a White Paper on Artificial Intelligence (see EUROPE 12408/19). All these documents are due to be published on 19 February. The White Paper on AI represents a first reflection and will be followed by “actions in the 4th quarter”, confirmed Mr Breton.
During his presentation, the Commissioner stressed three elements: - the importance of “do not over-regulate at the implementation”, because “it's too early”; - give the citizen freedom of choice so that he is not trapped by a platform or an algorithm; - control of our information systems and their security (legal, ethical and in cyberspace).
MEPs' position on automated decision-making
For their part, MEPs will be asked on 12 February to decide their position on automated decision-making, a concept that includes but is not limited to artificial intelligence, stressed Karen Melchior (Renew Europe, Denmark).
The document, validated by a large majority in the Committee on the Internal Market (IMCO), advocates a risk-based approach and raises the question of the competences of supervisory authorities with regard to the transparency and explicability of decision-making algorithms (see EUROPE 12410/14).
The Greens/EFA group is the only group to have tabled amendments, being mainly geared towards the sustainability of decision-making systems. The Greens/EFA also propose to opt for the reversal of the burden of proof in case of lack of conformity.
Allocation of tasks in the Committee on Legal Affairs
On the margins of the plenary session, the Committee on Legal Affairs presented to the press on Tuesday 11 February the three reports it intends to draw up on AI: a report prepared by Ibán García del Blanco (S&D, Spain) on the ethical aspects of AI, robotics and related technologies, a report on civil liability regimes for AI prepared by Axel Voss (EPP, Germany) and another on intellectual property rights for the development of Artificial Intelligence technologies prepared by Stéphane Sejourné (Renew Europe, France).
It should be noted that on the same day, the Renew Europe group held a press conference to present its position on AI. In a 30-page document, the group notes, for example, that authorities “should avoid using surveillance technologies using AI (such as facial recognition software) before the results of impact assessments are available and, if necessary, before legislation governing the use of these technologies is adopted”. [Draft resolution: https://bit.ly/2Hgrm12 ] (Original version in French by Sophie Petitjean, with Agathe Cherki)