Europe, US disagree on AI policy as EU AI Act awaits agreement in Brussels

Photo of author


Policymakers in Brussels remained locked in talks late Wednesday to reach an agreement on the world’s most ambitious law to regulate artificial intelligence. Meanwhile, senators indicated that the US Congress is taking a different stance from the EU on emerging technology, with lawmakers raising concerns that the bloc’s approach could be heavy-handed and risk alienating AI developers .

The split-screen incidents on both sides of the Atlantic underline the challenges of regulating artificial intelligence, which is a growing priority for governments around the world after the release of ChatGPIT, an AI-powered chatbot, caused a global frenzy.

Congress lags far behind its counterparts in Brussels, where a framework for regulating AI was first proposed in 2021. But after years of work, the future of the EU’s AI law remains uncertain amid lobbying and opposition from France, the EU’s largest country. Germany and Italy.

Schumer launches ‘all hands on deck’ push to regulate AI

After more than half a year working on AI policy, Senate Majority Leader Charles E. Schumer (D-N.Y.) told reporters that the bipartisan group is “really starting to work on the legislation,” though he was tight-lipped about it. I gave some specific information that will be included in such a bill.

The comments came during the last of two congressional AI forums in 2023, where lawmakers met with top tech executives, including former Google CEO Eric Schmidt, to better understand topics including AI doomsday and national security risks.

Senator Mike Rounds (R-S.D.), a member of the bipartisan working group that Schumer assembled to craft AI policy, said the senator would seek an “incentive-based approach” in an effort to retain AI developers in the United States. “Taking an approach.

“If [European policymakers] “Look at this as a regulatory activity, they will pursue AI development in the United States,” he told reporters after the two forums. “We don’t want to pursue AI development for our adversaries.”

Europe moves forward on AI regulation, challenging the power of tech giants

Meanwhile, EU officials sought late action on an EU AI Act, which would take a largely “risk-based” approach to limiting the use of AI applications, based on whether lawmakers would consider it appropriate. How dangerous they can be.

European Parliament representatives are struggling to counter efforts by the largest countries in the 27-member bloc to weaken the landmark bill. In recent weeks, negotiations between the EU’s various bodies – the European Commission, which proposes laws, and the European Council and the European Parliament, which adopt them – have been plagued by divisions that have hindered the creation of an Act. Years have been put at risk. Officials joined the talks with the hope that a compromise could be reached, and by midnight in Brussels negotiations were still ongoing.

If no agreement is reached in marathon deal-making that lasted until Thursday morning Brussels time, negotiations will likely move to a final effort in January, after which experts say any legislation will be unlikely to be passed before European Parliament elections. It may be difficult to do. In June.

“If we go beyond January, I think we are lost,” said Brando Benfi, one of two lawmakers taking the lead on the legislation in the European Parliament. “It will take at least nine more months for us to get the AI ​​Act.”

The EU’s largest countries have sought to remove a part of the bill that would impose binding regulations and transparency rules on foundation models, such as ChatGPT’s underlying technology, which generates answers based on models trained by scraping data from the Internet. Does it. Arguing that those regulations could stifle innovation and cause Europe to fall behind the United States in the race to develop such models, those countries were pushing for industry self-regulation instead.

People familiar with the talks, describing the delicate negotiations on condition of anonymity, said France appeared to be the strongest obstacle to a deal, in part to protect Paris-based Mistral, a growing company that develops AI foundation models. Based on his wishes. As well as other French AI firms. Meanwhile, efforts to limit AI in police work come as France prepares to deploy AI-powered smart cameras for policing and security at the 2024 Summer Olympics and French cities are already using such technology. Have entered legal gray areas by deploying or testing.

When asked about French opposition, France’s digital minister, Jean-Noël Barrot, said that European governments broadly oppose restrictions on the use of AI for policing and national security, and that strict rules on foundation model developers would be a major threat to the European Union. Can seriously hinder innovation.

“There is consensus within the Council that the use of AI for national security purposes should not be included in the regulation,” he said.

He further said, “The [AI] The industry in Europe has expressed its concern that placing too much of the burden on the shoulders of basic model developers is tantamount to those models not being developed in Europe,” he said.

Barrot stressed that the kind of compromise the French are seeking would still result in the world’s strongest legislation governing AI. He described the bill as the beginning as opposed to the end of European regulations on technology.

“I challenge anyone to introduce regulation that is as stringent as the EU AI Act worldwide,” he said.

Heading into Wednesday’s negotiating session, Benfi said that pressure by France and other countries to allow the industry to self-regulate would eliminate one of the most important elements of the bill, arguing that the agreement imposing a de facto ban would Must be discovered.

“The most powerful models will form the basis of all AI,” he said. “If we regulate their security and the transparency of how they work and the data used to train them, we will make it safer for all AI systems in the chain.”

Leave a comment