Europe agrees historic deal on AI regulation By Reuters

By Foo Yun Chee, Martin Coulter and Supantha Mukherjee

BRUSSELS/LONDON/STOCKHOLM (Reuters) – Europe reached a provisional agreement on Friday on landmark European Union rules governing the use of artificial intelligence, including how governments use ai in biometric surveillance and how to regulate systems of ai like ChatGPT.

With the political agreement, the EU moves towards becoming the first major world power to enact laws governing ai. Friday's agreement between EU countries and members of the European Parliament came after nearly 15 hours of negotiations that followed a nearly 24-hour debate the day before.

The two sides will discuss details in the coming days, which could change the shape of the final legislation.

“Europe has positioned itself as a pioneer, understanding the importance of its role as a global standard setter. I think this is a historic day,” European Commissioner Thierry Breton said at a press conference.

The agreement requires basic models such as ChatGPT and general purpose artificial intelligence (GPAI) systems to meet transparency obligations before coming to market. These include the production of technical documentation, compliance with EU copyright law and the dissemination of detailed summaries on the content used for training.

Basic high-impact models with systemic risk will have to perform model assessments, assess and mitigate systemic risks, perform adverse testing, report serious incidents to the European Commission, ensure cybersecurity and report on their energy efficiency.

GPAIs with systemic risk may rely on codes of practice to comply with new regulation.

Governments can only use real-time biometric surveillance in public spaces in cases of victims of certain crimes, prevention of genuine, present or foreseeable threats, such as terrorist attacks, and search for people suspected of the most serious crimes.

The agreement prohibits cognitive behavioral manipulation, non-selective removal of facial images from the Internet or CCTV images, social scoring and biometric categorization systems to infer political, religious, philosophical beliefs, sexual orientation and race.

Consumers would have the right to lodge complaints and receive meaningful explanations, while fines for violations would range from €7.5 million ($8.1 million) or 1.5% of global turnover to €35 million or 7% of global turnover.

Business group DigitalEurope criticized the rules as an additional burden on businesses, on top of other recent laws.

“We have a deal, but at what cost? We fully support a risk-based approach based on the uses of ai, not the technology itself, but the last-minute attempt to regulate base models has turned this on its head. “said CEO Cecilia Bonefeld-Dahl.

Privacy rights group European Digital Rights was equally critical.

“It's hard to get excited about a law that, for the first time in the EU, has taken steps to legalize live public facial recognition across the bloc,” said its chief political adviser, Ella Jakubowska.

“While Parliament fought hard to limit the damage, the overall package on biometric surveillance and profiling is lukewarm at best.”

The legislation is expected to come into force early next year once both sides formally ratify it and should apply two years later.

Governments around the world are trying to balance the advantages of technology, which can engage in human-like conversations, answer questions and write computer code, with the need to establish security barriers.

Europe's ambitious ai rules come as companies like OpenAI, in which Microsoft (NASDAQ:) is an investor, continue to discover new uses for their technology, prompting both applause and concern. Google owner Alphabet (NASDAQ ) on Thursday launched a new artificial intelligence model, Gemini, to rival OpenAI.

The EU law could become the model for other governments and an alternative to the lax approach of the United States and China's interim rules.

(1 dollar = 0.9293 euros)