London
CNN
–
Amazon and Meta executives told CNN this week that some concerns about artificial intelligence are over and that the European Union's new AI rules Risk of stifling innovation.
The EU gave the final green light to its AI Act on Tuesday, the same week tech leaders gathered in Paris for the annual VivaTech conference.
The first-of-its-kind law is set to reshape how firms and other organizations in Europe use AI for everything from healthcare decisions to policing. It imposes a blanket ban on using the technology in ways it deems “unacceptable” — for example, for social scoring.
The regulation also creates new disclosure obligations for large AI companies and requires greater transparency on uses of AI that are considered “high risk,” such as for education and employment.
For Yann LeCun, head of META AI, the “big question” about the new law is “should research and development in AI be regulated?”
“There are provisions in the EU AI Act and various other places that regulate research and development. I don't think it's a good idea,” he told CNN's Anna Stewart at the Paris event.
LeCun, widely regarded as one of the “godfathers of AI,” disagrees with concerns that AI may soon surpass human intelligence.
“I'm not sure it's anywhere close,” he said. “I don't think it's that dangerous, frankly, certainly not today.”
He added that AI systems may become much smarter in the future but then they will be designed with adequate safeguards.
“But today, trying to figure out how to make future super-intelligent AI systems safe is like asking in 1925, 'How do we make jet transport safe?' And jet transport hadn't been invented yet,” he said.
Amazon ( AMZN ) chief technology officer Werner Vogels echoed concerns that AI regulation could stifle innovation in some sectors.
When thinking about risks, regulators should consider the application of new technologies, For example, healthcare and financial services differ in its use. To summarize the meetings, he told Stewart.
“There are a lot of areas where I think the risks are low and we should let the innovation drive there,” he said. In other areas, where errors can have a greater impact on people's lives, risk management must be “unique to that particular area”.
Vogels warned the EU against over-regulating AI, pointing to the example of its signature data privacy law, called GDPR, which he described as a very “thick” book.
“Let's make sure that the regulatory requirements that we put in place, companies – not just the biggest companies but every company in Europe – can actually implement,” he said.
“We need to make sure that innovation continues and that innovation doesn't just come out of Europe. We already have a very long history of underinvestment in R&D in Europe.