Gavin Newsom Vetoes AI Safety Bill, Weighs Innovation and Regulation

Governor Gavin Newsom has vetoed a landmark California bill that aimed to introduce safety vetting requirements for powerful AI models. This decision aligns him with Silicon Valley tech giants and key congressional Democrats, marking one of the most significant legislative battles of the year.

In his veto message, Newsom expressed concerns about the bill's approach, stating that it “does not take into account whether an AI system is deployed in high-risk environments, involves critical decision-making, or the use of sensitive data.” Instead, he argued that applying stringent standards to even the most basic AI functions—relying on companies to deploy these measures voluntarily—was not the best approach to protect the public from real risks posed by AI technologies.

California’s Potential National AI Standards and Fierce Lobbying

This bill attracted significant attention due to its potential to set a national precedent for AI technology regulation. Newsom faced heavy lobbying from a wide range of influential California stakeholders, including Hollywood actors, major tech companies, investors, and prominent congressional members like former Speaker Nancy Pelosi.

The governor has long cautioned against hindering California's thriving AI industry, emphasizing the importance of maintaining the state’s economic competitiveness. However, he has also consistently advocated for California to lead in establishing responsible AI regulations.

Newsom’s Alternative Approach to AI Regulation

Though Newsom vetoed the sweeping legislation, he signed a more moderate bill requiring California’s emergency response agency to study AI risks. He also committed to developing AI guardrails in collaboration with academic leaders, such as Stanford's Dr. Fei-Fei Li, who opposed the original bill.

In addition, Newsom promised to work with organized labor and private businesses to expand AI applications in the workplace, building on pilot projects already underway in California state agencies.

The Divisive AI Bill and Its Proponents

The bill, introduced by state Sen. Scott Wiener, would have mandated that the largest AI models undergo safety testing before deployment, with the goal of preventing potential risks like the creation of bioweapons. Proponents argued that California needed to act swiftly to regulate AI technology, especially as Congress has been slow to address this issue.

Wiener expressed disappointment in Newsom's veto, claiming it was a missed opportunity for California to lead on innovative tech regulation. “We are all less safe as a result,” he said.

Divisions in Silicon Valley and Political Circles

The proposed bill divided Silicon Valley, with some prominent figures like Elon Musk supporting it as a step toward mitigating potential public risks. However, major tech companies like Google and OpenAI argued that the requirements would place an excessive burden on developers, particularly small startups.

Venture capital firms also rallied against the bill, hiring lobbyists to influence lawmakers. Wiener faced opposition from fellow Democrats representing Silicon Valley districts, including Pelosi and Rep. Ro Khanna, as well as local leaders like San Francisco Mayor London Breed.

Newsom’s Close Ties with Silicon Valley

As a former San Francisco mayor, Newsom has long maintained strong connections with Silicon Valley. His administration has worked to integrate AI into California's state government, including partnerships with Nvidia to establish AI training programs and Google to fund local journalism.

Legislative Year Dominated by AI Concerns

Concerns about the impact of AI dominated California’s legislative agenda this year. Lawmakers introduced several bills aimed at regulating AI's influence on elections, worker rights, and public safety. However, tech companies and business groups, such as the California Chamber of Commerce, resisted these efforts, arguing they would stifle innovation and infringe on free speech.

Despite the veto, Newsom did deliver some victories for advocates of stricter AI regulations, including signing laws that limit the use of digital likenesses in entertainment and combatting “deepfakes” used to impersonate political candidates.

The Path Forward for AI Regulation

Newsom's veto reflects the challenge of balancing innovation with regulation, as California continues to foster its position as a leader in AI development. While this decision delays immediate safety measures, it leaves room for more targeted and collaborative regulation in the future—ensuring public safety without unnecessarily hindering technological growth.