OpenAI opposes California AI bill

(ANSA) - ROME, AUG 23 - OpenAI, the start-up founded by Sam Altman that launched ChatGPT, is opposing a Californian bill - the so-called SB 1047 - aimed at averting risks from advanced AI models.
    In a letter addressed to the bill's sponsor, Scott Wiener, a California state senator, Jason Kwon, the start-up's chief strategy officer, warned against the fact that the upcoming bill represents a threat to "California's unique status as world leader in artificial intelligence".
    The bill, he wrote, could "slow down the pace of innovation and lead world-class Californian engineers and entrepreneurs to leave the State in search of greater opportunities elsewhere".
    OpenAI is the latest Silicon Valley company to voice doubts on California's attempt to regulate artificial intelligence in order to mitigate the risks linked to the new technology and its future developments. For his part, Wiener has defended the bill, which is expected to be voted on by the end of the month, criticizing the "tired argument" that tech start-ups would move away if the measure were approved.
    The bill foresees an obligation for AI companies to test their systems and to set up additional security measures to avert them being potentially manipulated, for example to destroy the State's electricity grid or to help build chemical weapons, both scenarios deemed by experts plausible in the future. The obligation exists for systems that cost more than 100 million dollars in calculation power to be trained, a requirement that has so far not been met by any existing AI model. (ANSA).