MEPs warn EU against weakening landmark AI rules
As EU officials debate whether to ease obligations for AI companies
A group of key architects behind the EU’s Artificial Intelligence Act have urged the European Commission to resist “dangerous” attempts to dilute the legislation.
The warning comes amid discussions in Brussels that could see certain provisions, designed to regulate powerful AI models, become voluntary rather than mandatory. The move, if enacted, could exempt major US tech companies such as OpenAI and Google from stricter compliance requirements.
According to the Financial Times, EU officials are debating whether to ease obligations for AI companies, particularly around measures aimed at preventing the spread of violent and false content, as well as interference in democratic processes.
The discussions follow intense lobbying from both Big Tech and figures such as Donald Trump, who have challenged Europe’s stringent approach to AI regulation.
In a letter to the European Commission’s digital chief, Henna Virkkunen, several prominent members of the European Parliament expressed concerns that such concessions would introduce legal uncertainty and undermine democratic values. The letter - signed by key MEPs involved in drafting the AI Act, including Spain’s former minister for digitalisation and AI Carme Artigas - warns that weakening oversight could have serious economic and political repercussions.
“If providers of the most impactful general-purpose AI models were to adopt more extreme political positions, implement policies that undermine model reliability, facilitate foreign interference or election manipulation, contribute to discrimination, restrict the freedom of information or disseminate illegal content, the consequences could deeply disrupt Europe’s economy and democracy,” the MEPs wrote.
The AI Act, hailed as the world’s most comprehensive AI regulation, classifies AI systems into risk categories, with stricter requirements for models deemed high-risk, such as those used in healthcare or transport.
In addition, general-purpose AI models like Google’s Gemini, Meta’s Llama and OpenAI’s GPT-4 are subject to enhanced transparency obligations regarding their training data and potential biases.
At the centre of the current debate is the development of a “code of practice,” which aims to provide AI companies with guidelines for compliance. This document, being drafted by a panel of experts including Turing Award winner Yoshua Bengio, is expected to be finalised in May. However, striking a balance between regulatory enforcement and industry cooperation has proven challenging.
USA hits out at Europe
Brussels has faced significant pushback from US companies over the AI Act. Meta’s head of global affairs, Joel Kaplan (yes, that Joel Kaplan), warned at a Brussels event in February that the proposed code of practice imposes “unworkable and technically unfeasible requirements.” The company has since announced it will not launch certain AI products in the EU due to the bloc’s strict privacy regulations.
Other major firms, including Google, Spotify and Ericsson, have also voiced concerns over the legislation’s impact on innovation.
Meanwhile, US Vice President JD Vance criticised Europe’s regulatory approach to AI during a speech at France’s AI Summit last month, arguing against “excessive regulation” and asserting that “AI must remain free from ideological bias.”
The new European Commission, which took office in December, has signalled a shift towards attracting AI investment rather than imposing additional regulatory burdens. As part of this broader push for deregulation, the commission recently scrapped plans for an AI liability directive.
Despite these changes, Virkkunen reaffirmed the EU’s commitment to maintaining core regulatory principles. Speaking at a Financial Times event on Tuesday, she said the code of practice should “help industries, stakeholders and SMEs” rather than introduce additional barriers.
Virkkunen also emphasised that the EU remains dedicated to ensuring a “fair, safe and democratic digital environment.”