Don't drop AI liability mechanism, lead lawmaker warns Commission

The European Parliament's lead lawmaker responsible for steering new AI Liability rules - German European People's Party MEP Axel Voss - has slammed a European Commission plan to withdraw its proposal as a “strategic mistake”.

The Commission’s 2025 work program will be presented in Strasbourg on Wednesday, setting out timelines for legislative initiatives in the coming year, but also flagging which proposals will be withdrawn. An AI Liability Directive was proposed in parallel with the Bloc’s AI Act in 2022 but according to the EU executive “no foreseeable agreement” is seen on the proposal in the coming year.

The Commission proposed the AI Liability Directive in 2022 in a bid to modernise existing rules with new provisions covering harms caused by AI systems and to ensure uniformity of protection.  

The document is now on the list of withdrawals. “The Commission will assess whether another proposal should be tabled or another type of approach should be chosen,” the text says.

Voss, who this month began a consultation to gather industry reactions on the scope of the rules to prepare the work in Parliament, told Euronews that the scrapping of the rules will mean “legal uncertainty, corporate power imbalances, and a Wild West approach to AI liability that benefits only Big Tech."

"The reality now is that AI liability will be dictated by a fragmented patchwork of 27 different national legal systems, suffocating European AI startups and SMEs,” he added.

Scepticism

The Brussels tech lobby and consumer organisations were sceptical about the need for additional rules. They claim that the issues are already covered under the revamped Product Liability Directive (PLD). 

A study by the Parliament’s research service, presented in the Parliament’s legal affairs committee (JURI) late January, pointed to problems that might arise as a result of large language models, including ChatGPT and Claude.ai, falling out of the scope of the current Product Liability Directive.

Voss said that unlike the AI Act, which is an up-front regulatory framework, the liability directive would only apply when AI actually caused harm.

"With such strategic mistakes, I unfortunately lack the confidence that we will ever be able to create a future-oriented digital single market," Voss said.