Brussels, under American pressure, is withdrawing one of the three pillars of its legislation on artificial intelligence (AI) that it intended to put in place. On the evening of Tuesday 11 February, the closing day of the Paris AI Summit and the scathing speech by US Vice-President V.D Vance, the European Commission published its work programme for 2025, with the list of regulatory projects that it intended to advance, or on the contrary put aside.
Among the projects that were deleted: the future directive, presented in 2022, which was to resolve the issue of liability in the event of physical or moral damage caused to European citizens by AI tools such as ChatGPT (OpenAI) or Le Chat (Mistral).
A text that did not meet with unanimous approval, explains the Commission
But that’s over now. The second text, the draft directive on AI liability, has disappeared from the EU executive’s 2025 work programme. In the EU system, the European Commission has the legislative initiative – it is the Commission that can propose or withdraw draft EU legislation. Brussels explains this deletion by a “lack of foreseeable agreement”.It adds that it will “assess whether to present another proposal or to opt for a different type of approach” on this subject.
It is true that the draft directive was far from unanimous, whether within the 27 Member States or the European Parliament. While some felt that the text would complicate existing legislation, others thought on the contrary that the draft directive did not go far enough. The latter argued for the text to clearly introduce strict liability for certain AI applications.
Within the European Commission, Henna Virkkunen, the EU's head of digital policy, is said to have used all her weight to have the text deleted in the name of regulatory simplification, reports MLex. Michael McGrath, the European Commissioner for Justice, is said to have only weakly defended the proposal, our colleagues explain. The European Commission has been trying for several weeks to simplify the regulatory framework and the administrative burden on companies. But the deletion of the Commission’s work programme remains a surprise: discussions and public consultations on the subject had resumed in recent weeks, after several breaks.
European citizens, the big losers from this deletion?
Axell Voss, the German MEP (European People’s Party) rapporteur of the text, wrote in a LinkedIn message that “by deleting the AI liability directive, the Commission is actively opting for legal uncertainty, an imbalance of power between companies and a “Wild West” approach to AI”.
For the rights association Centre for Democracy and Technology (CDT Europe), this decision is above all bad news for European citizens.
The AI Liability Directive was intended to facilitate legal recourse for Europeans "who seek justice when they are harmed by an AI system" 'Segoe UI Symbol', 'Noto Color Emoji';">". «The harms caused by AI systems and models are notoriously difficult to prove, due to their complexity and lack of transparency», recalls Laura Lazaro Cabrera, legal advisor to CDT Europe, quoted in the association's press release published on February 12.
With the directive thrown into oblivion, individuals will now have "limited means of seeking redress when they suffer harm caused by AI", she laments. Namely: the use of the 27 national laws on civil liability (of the EU), not necessarily adapted to the issues of AI tools.
"We should respond firmly to a Trump government that is walking all over us"
For her part, French MEP Leïla Chaibi, who speaks on her BlueSky account, deplores this turnaround which can be explained by "the attacks of American Vice President J.D. Vance". A few hours earlier, the American Vice President gave a virulent speech at the podium of the Paris AI Summit, criticizing the DSA and the GDPR, regulations described as excessive and onerous. "America will not accept" legislation that attempts to "tighten the screws" on American AI champions, he hammered home.
"We should respond firmly to a Trump government that is treading on us. Von der Leyen's EU prefers to play the docile auxiliaries", laments the MEP (the Left) on her account.
For others, the withdrawal is mainly intended to show the new pro-competitive and pro-innovation approach of the European Commission. The removal of the directive would show the goodwill of the executive towards the new American administration. Donald Trump advocates deregulation, never missing an opportunity to criticize European regulation on digital technology and artificial intelligence.
0 Comments