Gavin Newsom has caved to the billionaires

SB 1047's veto is a win for tech companies — and a loss for everyone else

In the annals of technology regulation, California's Senate Bill 1047 may one day be seen as a missed opportunity of historic proportions.

For a fleeting moment, it seemed politicians might rise to the challenge of artificial intelligence. SB 1047, an ambitious effort to regulate AI, sailed through the state legislature despite fierce opposition from tech billionaires and corporate interests. It united an unlikely coalition: activists, scientists, AI workers, actors, unions, and the general public — all urging Governor Gavin Newsom to sign the bill into law.

But in a depressingly predictable turn of events, the people lost the fight. Newsom, ignoring his constituents, caved to pressure from Silicon Valley’s elite. His veto of SB 1047 leaves some of AI's most concerning risks unaddressed — and marks yet another victory for gigantic corporations in evading government oversight.

At its core, SB 1047 was eminently reasonable. It would have required the largest AI companies to test their models for dangerous capabilities (such as the ability to cause large-scale cyberattacks, or help terrorists develop biological weapons), and take steps to mitigate those risks.

Notably, companies like OpenAI, Google, and Meta have already promised to do this, under commitments made to the White House and other governments. The bill would have simply made these promises binding, replacing hollow assurances with genuine accountability. And in the event something did go wrong and companies hadn’t taken “reasonable care” to prevent the harm, they could be found liable for the damages caused. As employees of leading AI companies told Newsom, the bill’s requirements were both “feasible and appropriate”.

The campaign against the bill, however, painted a vastly different picture: because it relied on outright lies. Andreessen Horowitz, the venture capital giant which led the campaign against the bill, made multiple false claims about its content. Y Combinator falsely claimed that developers could face jail time for “failing to anticipate misuse of their software” — a claim that bill author Scott Wiener called “categorically false”. Even AI luminary Fei-Fei Li joined in, falsely arguing that one of the bill’s provisions would “devastate the open-source community”, despite open-source technology having explicit carve outs from that provision. 

Incredibly, Democratic officials amplified these distortions. Representatives Zoe Lofgren, Anna Eshoo, Ro Khanna and others claimed the bill required adherence to non-existent guidance — despite that guidance existing (and the bill not requiring adherence to it). Nancy Pelosi, meanwhile, called the bill “ill informed”, ignoring the endorsements of many prominent AI experts, including two of the “godfathers of AI”.

These lawmakers insist that federal regulation is the answer, conveniently ignoring political reality: with staunch Republican opposition, federal legislation is a pipe dream. SB 1047 represented the only viable path to AI regulation in the near term — yet Democratic elites torpedoed it.

Their motivations are transparent: Silicon Valley still holds enormous sway over the Democratic Party. Google, a major donor to Lofgren, Eshoo, and Khanna, lobbied hard against the bill. Andreessen Horowitz, despite its far-right leanings, hired Newsom confidant Jason Kinney as a lobbyist — and appears to have the ear of Lofgren. Then there's Ron Conway, a Democratic mega-donor close to Pelosi and Newsom, who owns stakes in OpenAI, Anthropic, and Mistral. Conway reportedly lobbied hard to kill the bill, seemingly threatening to ruin Wiener’s career over it.

With presidential ambitions looming, Newsom evidently found alienating these power brokers too politically costly. But in vetoing SB 1047, he may have made a grave miscalculation. Polling shows overwhelming public support for AI regulation, both in California and nationwide. A growing coalition of advocacy groups, including the Screen Actors Guild, had thrown their weight behind the bill. Voters have indicated they would hold Newsom personally responsible in the event of a future AI-related catastrophe. They’d be right to do so.

Newsom has suggested that he vetoed 1047 because he actually wants more expansive legislation. That logic is nonsensical: such legislation has even less chance of passing. Voters would do well to ignore this transparent attempt at misdirection and see the reality: Newsom has blocked America’s best shot at regulating AI, prioritizing billionaires’ interests over public safety.

Gavin Newsom’s failure of leadership may come to be seen as a pivotal missed opportunity — a moment when we could have established guardrails for a transformative technology, but chose instead to let it run wild. The governor has made his choice clear: he stands with the billionaires, not the people. But at least those billionaires are happy.