New York State Legislators Passed a bill Thursdayauda this aims to prevent national AI models from Openai, Google and Antropic to contribute to catastrophic scenes, including the death or injury of more than 100 people, or more than $ 1 billion in damage.
The passage of the lever law represents a victory for the AI -Security Movement that has lost land in recent years As Silicon Valley and the Trump administration prioritized speed and innovation. Security advocates including Nobel laureate Laureate Geoffrey Hinton and Ai Research Pioneer Yoshua Bengio championed the Lela Law. If it becomes a law, the bill would establish America’s first set by legally ordered transparent standards for Frontier Ai Labs.
The LEA law has many same supplies and goals as California’s controversial AI safety bill, SB 1047, who was ultimately vetoed. However, the bill’s co-sponsor, New York State Senator Andrew Gounardes told Techcrunch in an interview that he deliberately designed the law law such that it does not cool innovation between startups or academic researchers-frequent criticism of SB 1047.
“The window to replace guards is rapidly declining, considering how quickly this technology is evolving,” said Senator Gounardes. “The people who know [AI] The best says these risks are incredibly likely […] That’s alarming. ”
The Law Act is now directed to the desk of the Governor Kathy Hochul, where it could sign the law, send it for amendments or veto it at all.
If signed in law, New York’s AI security bill would require the world’s largest AI labs to publish thorough reports on security and security on their Frontier AI models. The bill also requires that AI labs report on security incidents, such as AI model -behavior or bad actors stealing AI model if they occur. If TE Companiesnological Companies do not meet these standards, the LA law authorizes the New York prosecutor to bring civil penalties of up to $ 30 million.
The LEA law aims to closely regulate the world’s largest companies – whether they are based in California (such as Openai and Google) or China (such as Deepseek and Alibaba). The transparent requirements of the bill apply to companies, whose AI models have been trained using more than $ 100 million in computer resources, and are available to New York residents.
Silicon Valley repulsed significantly on the AI security bill of New York, the New York State Assembly and co-sponsor of the LEA Law Act Alex Bores told Techcrunch. Bores called the industrial resistance not surprising, but asserted that the elevation law would not limit TE companiesnological companies in any way.
Anthropical, the security focus AI lab that called Federal Transparent Standards for AI Companies Earlier this month, did not reach an official stance on the bill, co -founder Jack Clark said in Friday post at X. However, Clark expressed some complaints about how wide the LEA law is, noting that it could present a risk to “smaller companies.”
When asked about Antropic’s criticism, State Senator Gounardes told Techcrunch that he thought it was “missing the brand”, noting that he had designed the bill not to apply for small companies.
Openai, Google and Meta did not respond to Techcrunch’s request for comment.
Another common criticism of the LEA law is that AI model developers simply would not offer their most advanced AI models in the state of New York. This was a similar criticism of SB 1047, and it is mostly played in Europe thanks to the difficult regulations on technology on the continent.
Assembly Bores told Techcrunch that the regulatory burden of the LEA law is relatively light, and therefore should not require that technology companies stop operating their products in New York. Given that New York has the third largest GDP in the United States, drawing out of the state there is not something that most companies would take lightly.
“I do not want to underestimate the political courtesy that could happen, but I am very confident that there are no economic reasons for them not to provide their models in New York,” said the Borres Assembly.
