OpenAI boss warns of ‘strict regulation’ of AI

(Tel Aviv) OpenAI boss Sam Altman spoke out on Monday against “strict regulation” of artificial intelligence (AI), which risks hampering its development, but stressing the need for oversight long-term institutional.

Last month, Mr. Altman, whose company created the ChatGPT interface, told a US parliamentary committee that government intervention to regulate AI will be “crucial” to “limit the risks” associated with this technology. technology.

The OpenAI boss, speaking at Tel Aviv University on Monday, stressed, however, that his call for AI oversight was not about “today’s systems.”

“I think it would be a mistake to strictly regulate this area or try to slow down this incredible innovation,” he said.

But the risk of misused “superintelligence” is real, he acknowledged. It’s “something we may have to grapple with over the next decade and it doesn’t give the institutions of the world much time to adapt. »

He again proposed the creation of an “international organization”, which would use “computing power and technical (tools)” to create “a framework […] licensing artificial intelligence interfaces, to verify their security system and offer (them) tests to pass. »

“It would be a way to deal with the risk” of having artificial intelligences diverted from their initial use, he said, deeming this risk “very serious”.

“We are doing the same for nuclear,” he said, citing the example of the International Atomic Energy Agency (IAEA).

In a meeting with Israeli President Isaac Herzog earlier in the day, Altman stressed “the importance and urgency of finding ways to limit these enormous risks.” “Everyone wants to find a solution,” he said.

Israeli Prime Minister Binyamin Netanyahu said Altman told him in a phone interview that Israel “could become a major global player in the field” of AI.

In a statement, Mr Netanyahu said his country – where the high-tech sector is already a heavyweight in the economy – must “develop a national policy” on AI.

Mr. Altman’s visit to Israel is part of an international tour aimed at reassuring people about the use of OpenAI by the general public and businesses.

OpenAI’s generative AIs, ChatGPT for text and Dall-E for images, launched to the general public at the end of 2022 with dazzling success and in a legal vacuum, have triggered an avalanche of fears, particularly regarding misinformation. and possible job destruction.