California Gov. Gavin Newsom signs landmark bill creating AI safety measures

Sacramento, California – California governor Gavin Newsom signed a law on Monday that aims to prevent people from using powerful artificial intelligence models for potentially catastrophic activities such as the construction of a bio-arme weapon or the closure of a banking system.

This decision comes as Newsom presented California as a leader in AI regulation and criticized inaction at the federal level during a recent conversation with former President Bill Clinton. The new law will establish some of the first nations regulations on large -scale AI models without harming the local state industry, Newsom said. Many of the best IA companies in the world are located in California and will have to follow the requirements.

“California has proven that we can establish regulations to protect our communities while ensuring that the growing AI industry continues to prosper. This legislation establishes this balance,” Newsom said in a statement.

Legislation obliges companies to implement and disclose public security protocols to prevent their most advanced models from being used to cause major damage. The rules are designed to cover AI systems if they respect a “border” threshold which signals that they execute on a huge amount of calculation power.

These thresholds are based on the number of calculations that computers carry out. Those who have developed the regulations have recognized that digital thresholds are an imperfect starting point to distinguish the most efficient generative systems of today from the next generation which could be even more powerful. Existing systems are largely manufactured by California -based companies such as Anthropic, Google, Meta Platforms and Openai.

The legislation defines a catastrophic risk as something that would cause at least $ 1 billion in damage or more than 50 injuries or deaths. It is designed to guard against AI used for activities that could cause mass disturbances, such as hacking an electrical network.

Companies must also report to the State any critical security incident within 15 days. The law creates denunciation protections for AI workers and establishes a public cloud for researchers. It includes a fine of $ 1 million per violation.

He attracted the opposition of certain technological companies, which argued that AI law should be carried out at the federal level. But Anthropic said that regulations are “practical guarantees” that officially make security practices that many companies already do voluntarily.

“While federal standards remain essential to avoid a patchwork of state regulations, California has created a solid framework that balances public security with continuous innovation,” said Jack Clark, co-founder and head of politics at Anthropic, in a statement.

The signing comes after Newsom last year opposed a broader version of the legislation, moving with technological companies that declared that the requirements were too rigid and would have hampered innovation. Newsom has rather asked a group of several industry experts, including the Pioneer of the Fei-Fei Li, to develop recommendations on railing around powerful AI models.

The new law incorporates recommendations and comments from the group of AI Newsom and industry experts, supporters said. The legislation also does not put the same level of declaration requirements on startups to avoid harming innovation, said the Senator of the Scott Wiener State of San Francisco, the author of the bill.

“With this law, California intensifies, once again, as a world leader in technological innovation and security,” Wiener said in a statement.

Newsom’s decision comes as President Donald Trump announced in July a plan to eliminate what his administration considers “expensive” regulations to accelerate AI innovation and cement the position of the United States as a world leader in AI. Congress Republicans at the start of this year tried without success to ban states and localities from regulating AI for a decade.

Without stronger federal regulations, states across the country have spent the last years trying to slow down technology, by attacking everything, deep buttocks in the “therapy” of AI. In California, the Legislative Assembly has adopted a certain number of bills this year to respond to safety problems concerning AI chatbots for children and the use of AI in the workplace.

California has also been an early adopter of AI technologies. The state has deployed generative AI tools to identify forest fires and approach the congestion of roads and road safety, among others.

___

The journalist of Associated Press, Matt O’Brien, contributed to the report.