European corporations hate the EU’s new AI guidelines – this is why

Artificial intelligence is definitely keeping Swiss Post on its toes. After a spate of open letters about runaway AI, unregulated AI and apocalyptic AI, another letter arrived at the EU’s door today.

In this case, however, the signatories have expressed an opposite concern. Instead of calling for more rules, they fear there will soon be too many.

Their goal is the upcoming AI law. The new rules are considered the world’s first comprehensive legislation for this technology and seek to walk the fine line between ensuring safety and encouraging innovation. The new letter, signed by executives from some of Europe’s biggest companies, warns they are losing their balance.

“The draft law would jeopardize Europe’s competitiveness and technological sovereignty without effectively addressing the challenges we face and are facing,” reads the letter, which was sent to the European Parliament, the Commission and member states.

“States with the best performing models will have a distinct competitive advantage.

Signatories include top executives from corporate giants like Heineken, Carrefour and Renault, as well as executives from tech companies like Ubisoft, TomTom and Mistral AI. They say the AI ​​law will cause companies to leave the Union, divert cash from investors and slow down development in Europe.

The <3 of EU technology

The latest rumors from the EU tech scene, a story of our wise founder Boris and questionable AI art. It’s in your inbox for free every week. Join Now!

One of their main concerns stems from a recent rule change. On June 14, the European Parliament added new requirements for generative AI tools like ChatGPT, which signatories said would entail “disproportionate” compliance costs and liability risks.

They warn that this will cause Europe to fall further behind the US in AI development. That impact, they continue, will extend from business to culture as large language models become embedded in everything from search engines to digital assistants.

“States with the best-performing large language models will have a key competitive advantage… Europe cannot afford to remain on the sidelines,” the letter reads.

Bend AI rules

In addition to raising the grievances, business leaders proposed some solutions.

Her main proposal is to limit EU laws to general principles in a risk-based approach to be implemented by a dedicated body and adapted to new advances and risks. Such a process must be developed in dialogue with business.

The signatories also expressed their support for some aspects of the AI ​​law. In particular, they advocated mandatory security testing for new systems, standard labeling of AI-generated content, and due diligence in model development.

However, these attempts at overtures have not convinced the legislature. Dragos Tudorache, who helped draft the AI ​​law, promptly rejected the letter.

“I am convinced that they did not read the text carefully, but rather responded to the impulse of a few who have a vested interest in the subject,” he told Reuters.

It is positive for the authors that there is still time to write many more letters. The AI ​​law is not expected to come into force before 2026.

Comments are closed.