The European Parliament wants to participate in the development of the EU artificial intelligence strategy, which the European Commission will present on 19 February. On Thursday, 23 January, MEPs on the Committee on the Internal Market adopted a draft question and resolution that calls for a risk-based approach and requests additional powers for supervisory authorities.
All MEPs supported the question for oral answer from the Commission. Thirty-nine of them voted in favour of the draft resolution entitled “Automated decision-making processes: ensuring consumer protection and the free movement of goods and services”; four abstained. Both texts will be on the mid-February plenary session’s agenda.
The most ambitious provision of the draft resolution concerns the transparency and readability of algorithms. In addition to reiterating these objectives, the text “invites the European Commission to assess whether additional prerogatives should be granted to market surveillance authorities in this respect”. It also emphasises the need for unbiased algorithms and unbiased, high-quality data sets. Moreover, it calls for European citizens to be informed as to how an algorithm works, how decisions can be checked and corrected, and whether prices have been personalised. The project also calls for independent monitoring by qualified professionals when legitimate public interests are at stake.
Nevertheless, the text makes no reference to facial recognition, which the Commission is considering suspending for a fixed period, as it has provided for in the draft communication leaked to the press and explained in detail by EUROPE (see EUROPE 12406/8 and 12408/19). (Original version in French by Sophie Petitjean)