For months, Sam Altman, CEO of Microsoft-backed OpenAI has urged lawmakers around the world to draw up new rules governing the technology. On Wednesday, he threatened the ChatGPT maker may leave the EU if the bloc “overregulated”.
Altman has spent the past week crisscrossing Europe, meeting top politicians in France, Spain, Poland, Germany and the UK to discuss the future of AI, and progress of ChatGPT.
More than six months after OpenAI unveiled its AI-powered chatbot ChatGPT to the world, fears around its potential have provoked excitement and alarm – and brought it into conflict with regulators.
One place Altman didn’t get to this week was Brussels, where EU regulators are working on the long-awaited EU AI Act, which could be the first set of rules globally to govern AI.
Altman cancelled a scheduled visit to Brussels, two sources familiar with the matter said. OpenAI did not respond to a request for comment.
“The current draft of the EU AI Act would be over-regulating, but we have heard it’s going to get pulled back,” Altman said in London on Wednesday.
EU lawmakers responsible for shaping the AI Act disputed Altman’s claims. “I don’t see any dilution happening anytime soon,” Dragos Tudorache, a Romanian member of the European Parliament who is leading the drafting of EU proposals, told Reuters.
“We are nevertheless happy to invite Mr. Altman to Parliament so he can voice his concerns and hear European lawmakers’ thoughts on these issues,” he said.
EU industry chief Thierry Breton also criticised the threat, saying the draft rules are not for negotiation.
On Thursday, OpenAI is expected to discuss in more detail how AI should be regulated, amid Altman’s busy schedule of meetings with world leaders such as UK Prime Minister Rishi Sunak and French President Emmanuel Macron.
Lawmakers won’t be ‘blackmailed’
Dutch MEP Kim van Sparrentak, who has also worked on the draft EU law, said she and her colleagues “shouldn’t let ourselves be blackmailed by American companies.”
“If OpenAI can’t comply with basic data governance, transparency, safety and security requirements, then their systems aren’t fit for the European market,” she said.
By February, ChatGPT set a record for the fastest growing user base of any consumer application app in history.
OpenAI first clashed with regulators in March, when Italian data regulator Garante shut the app down domestically, accusing OpenAI of flouting European privacy rules. ChatGPT came back online after the company instituted new privacy measures for users.
Meanwhile, EU lawmakers added new proposals to the bloc’s AI Act, forcing any company using generative tools, like ChatGPT, to disclose any copyrighted material used to train its systems.
EU parliamentarians agreed on the draft of the act earlier this month. Member states, the European Commission and Parliament will thrash out the final details of the bill.
Through the Council of Europe, individual member states like France or Poland can also seek amendments before the bill is passed potentially later this year.
Plans in ‘full swing’
While the legislation has been in the works for several years, new provisions specifically targeting generative tools were drawn up only weeks ahead of a crunch vote on the proposals.
Reuters earlier reported some lawmakers had initially proposed banning copyrighted material being used to train generative AI models altogether, but this was abandoned in favour of stronger transparency requirements.
“These provisions relate mainly to transparency, which ensures the AI and the company building it are trustworthy. I don’t see a reason why any company would shy away from transparency,” Tudorache said.
Nils Rauer, a technology partner at law firm Pinsent Masons, said it was “no surprise” Altman had made his comments as lawmakers worked through their proposals.
“It is unlikely OpenAI will turn its back on Europe. The EU is economically too important,” he said. “You cannot carve out the single market, with close to 500 million people and a 15-trillion-euro ($16.51 trillion) economy.”
Altman was in Munich, Germany, on Thursday where he said he had met with Chancellor Olaf Scholz.
Sergey Lagodinsky, a German MEP who also worked on the legislation, said that while Altman may be trying to push his agenda among individual countries, Brussels’ plans to regulate the technology were “in full swing.”
“There may be some amendments, of course,” he said. “But I doubt they will change the overall trajectory.”