PARIS, FRANCE — The European Union on Thursday accused the boss of OpenAI, the firm that created the explosively popular ChatGPT bot, of “blackmail” by suggesting that new rules would force his company to leave the bloc.
Sam Altman told reporters in London that OpenAI could potentially “cease operating” in Europe if the bloc pushed ahead with its long-awaited AI Act, a regulation designed to protect the public from the technology.
Altman is in the middle of a global tour to charm leaders and powerbrokers from Lagos to London, but his comments appeared to anger EU industry commissioner Thierry Breton.
“There is no point in attempting blackmail — claiming that by crafting a clear framework, Europe is holding up the rollout of generative AI,” tweeted Breton, referring to the artificial intelligence software that generates original content.
ChatGPT burst onto the scene late last year, demonstrating an ability to generate essays, poems and conversations from the briefest of prompts.
Microsoft later laid out billions of dollars to support OpenAI and has begun using the firm’s technology in several of its products.
AI’s boosters claim the technology will improve lives by doing menial tasks better and revolutionising human interaction with machines.
But critics say it will decimate entire industries, lead to a flood of misinformation and copyright infringements, and entrench race and gender biases.
In comments first reported by Time magazine on Wednesday, Altman said his firm would “try to comply” with the EU’s regulation, which is unlikely to be operational until the end of 2025.
“If we can comply, we will, and if we can’t, we’ll cease operating,” he said, adding that there were “technical limits to what’s possible”.
The Financial Times newspaper carried similar quotes on Thursday, sparking the reply from Breton.
He said he was helping firms to deal with the rules, pointing to his idea for a “voluntary pact” on AI that would bridge the gap while lawmakers finalised the full regulation.
Microsoft’s President Brad Smith said during an event in the United States on Thursday that he was optimistic “reason will prevail” and that the final AI Act would be an acceptable compromise.