Kathy Hochul proposes major changes to New York's AI bill amid intense tech lobbying pressure

N.Y. Gov. Kathy Hochul proposes major changes to AI bill

Why it matters: States are increasingly the venue of intense tech lobbying as they get more active on regulating AI, at the same time that President Trump pushes ahead with plans to stop them with an executive order. * According to the sources who have reviewed the document, the original RAISE Act has been crossed out, with replacement text that nearly verbatim resembles a California AI safety bill that Gov. Gavin Newsom (D) signed this year, SB53. Driving the news: Hochul this week "redlined" the RAISE Act, an AI safety bill from NY state Sen. Andrew Gounardes and Assemblymember Alex Bores that was written to require incident reporting and safety plans for powerful AI models. The big picture: The debate has received national attention as pro-AI groups go after Bores, who is pushing for Hochul to sign the bill and is running for former Rep. Jerry Nadler's (D-N.Y.) seat in Congress. * This week, pro-AI PAC Leading the Future started running ads going after Bores in his district NY-12, saying the RAISE Act would "fail to keep people safe" and "crush jobs." * AI companies and startups have been urging major changes to the RAISE Act and suggesting it more closely mirror SB53. California's AI law that Hochul is looking to adopt requires large AI developers to make public disclosures about safety protocols and report safety incidents. * It also creates whistleblower protections and expands cloud computing access for smaller developers and researchers. What's inside: Per a person familiar with the negotiations, the RAISE Act no longer includes: * A requirement for detailed safety standards from AI companies that reasonably reduce the risk of harm. * A ban on releasing models that pose an unreasonable risk of harm. * A requirement to report security incidents; along with requirements that would apply to future models of foreign AI models like DeepSeek. The bill also has a weakened catastrophic risk disclosure component, the person said. A spokesperson for Hochul didn't immediately respond to a request for comment. * The news was first reported in Transformer. What's next: Hochul has ten days (not counting Sunday) from the time she is delivered bills to sign, veto or agree to "chapter amendments" with bill sponsors.

Gov. Hochul should kill Albany's toxic AI bill -- and let Congress...

Even though Congress has punted (for now) on any national rules to regulate Artificial Intelligence development, Gov. Kathy Hochul should still veto the bill now on her desk for New York-only regulation. America is absolutely guaranteed to lose the race with China to set the global standards for AI if US developers have to cope with 50 different sets of contradictory regulation -- and the nation would pay a huge long-term economic price for that loss. Hochul herself flagged that point months ago on Bloomberg TV: "People prefer to have a federal regulation," as "it's hard when one state has a set of rules, another state does, another state. I don't think that's a model for inspiring innovation." Beyond that, the bill passed by New York legislators, The Responsible Artificial Intelligence Safety and Education (RAISE) Act, contains some terrible provisions designed to boost trial lawyers, not cutting-edge tech. US companies want to make their designs open-source, allowing others to build on their tech so American software can become the global standard. But the ambulance chasers want to be able to sue one tech company for sky-high damages over how another programmer uses (abuses) such open-source code; that's tantamount to banning US open-source work -- even though US lawyers will have no ability to suck the blood out of China's open-source innovators. Similar innovation-crushing provisions are already part of the Transparency in Frontier AI Act that California passed last month; having New York second the trend is beyond ominous. At the very least, it's going to push cutting-edge tech companies to states like Texas and Florida, whose laws are likely to be far more innovation-friendly: On top of New York's high taxes and energy costs, lawsuit-friendly laws will make the Empire State a foolish place for tech to set up shop. Again, Congress really needs to act, with lawmakers meeting in the middle to set clear, practical AI guidelines. Among the musts for any national regulation, incidentally: Copyright must be protected, so that AI companies can't mine publishers' content without compensating them. That could put countless entertainment, news and media outfits, including us, out of business fast. But American tech companies can't be strangled, either, as the New York bill would do: It even tries to set international regulatory standards -- anyone who wants to do business in New York will have to abide by them: Another way to send more companies running full-speed from the Empire State. China's lack of any true rule of law (especially in protecting intellectual property) constitutes a real handicap in the AI race, but imposing actively bad laws in America could prove worse. Above all else, most folks who win New York state legislative races are terrible candidates when it comes to grasping the issues involved in cutting-edge tech; how can they possibly craft sensible policy that protects consumers while allowing innovation? Veto the RAISE act, governor, and stick to your guns: Shaping the regulatory future of AI is a job for national lawmakers.