news

한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina

it house reported on september 30 that california governor gavin newsom today vetoed the controversial frontier artificial intelligence model security innovation act (sb 1047 for short). he mentioned many things in his veto message. factors that influenced his decision,these include the burden the bill places on ai companies, california’s dominance in the field, and criticism that the bill may be too broad.

image source pexels

according to it house, the sb 1047 bill aims to establish security standards for developers whose training costs exceed us$100 million or reach a certain computing power to ensure the safe development of large-scale artificial intelligence models.the bill requires developers of large-scale, cutting-edge artificial intelligence models to take precautions such as pre-deployment testing, simulated hacking attacks, installing cybersecurity safeguards, and providing whistleblower protections.。

the bill was submitted to the california assembly in february this year and caused great controversy. on august 19, the california legislature announced a revised version of the bill. the bill was supported by jeffrey hinton and joshua bengio, known as the "godfathers of artificial intelligence," but was also opposed by scientists such as yang likun and li feifei, as well as leading artificial intelligence companies, saying that the law would "hindering innovation."

newsom said, “while sb 1047 has good intentions, it does not consider whether artificial intelligence systems are deployed in high-risk environments, involve critical decisions, or use sensitive data. on the contrary, as long as it is a large-scale system deployment, the bill even "i don't think that applying strict standards to the most basic functionality is the best way to protect the public from the real threats posed by this technology."

newsom also said the bill could "give the public a false sense of security about controlling this rapidly evolving technology," saying, "smaller, more specialized models could be just as dangerous or even more dangerous than the ones targeted by sb 1047 — — the price could be to undermine innovation that advances the public good.”

newsom said he agrees safety protocols and guardrails should be in place, with "clear and enforceable" consequences for bad actors. however, he noted that he believed the country should not “satisfy with a solution that is not informed by an analysis of empirical trajectories of ai systems and capabilities.”

jamie radice, public affairs manager at meta, said, “we are pleased that governor newsom has vetoed sb 1047. the bill would stifle ai innovation, hurt business growth and job creation, and break the state’s long tradition of promoting open source development. we support it. responsible ai regulations and remain committed to working with legislators to promote better approaches.”