Parliament Representative Calls for Preservation of AI Liability Directive
German MEP Axel Voss cautions that eliminating the AI liability regulations may result in legal ambiguities and an inconsistent regulatory environment throughout Europe.
The European Commission's decision to retract its proposed AI Liability Directive has been met with criticism from German MEP Axel Voss, a member of the European People's Party spearheading the push for new AI liability regulations in Parliament.
The Commission's 2025 work program, which is scheduled for presentation in Strasbourg on Wednesday, features a list of proposals slated for withdrawal, including the AI Liability Directive that was introduced in 2022 alongside the AI Act.
This directive was aimed at updating existing liability frameworks to tackle harms caused by artificial intelligence systems and to ensure consistent protection across the European Union.
The Commission's document states there is "no foreseeable agreement" on the proposal within the next year, leading to its planned withdrawal, though reassessment may occur later.
Voss, who initiated consultations this month to collect industry input on the scope of the proposed regulations, cautioned that scrapping the directive would create legal ambiguity and imbalance corporate power dynamics.
He emphasized that without a cohesive strategy, liability for AI-related damages would be dictated by a fragmented assortment of twenty-seven national legal systems, which could impede European AI startups as well as small and medium-sized enterprises.
The subject has also caused rifts among tech lobbyists and consumer advocacy groups in Brussels.
While industry advocates maintain that the modified Product Liability Directive already addresses many relevant concerns, consumer organizations have favored the enhanced protections suggested by the AI Liability Directive.
A report from the European Parliament's research service identified potential deficiencies in the current legal framework, particularly concerning large language models like ChatGPT and Claude.ai.
European Commissioner Maroš Šefčovič, who discussed the Commission's intentions at a press conference in Strasbourg, remarked that several proposals have been stalled for extensive periods, fostering skepticism about their advancement this year.
He mentioned that the withdrawal list is designed to stimulate discussion among co-legislators and that the Commission may reconsider its stance if adequate support is demonstrated for additional work on the directive in the coming year.