Newsom's AI kill switch builds a moat for corporate giants
By Caroline Ashford · Reporting from Richmond, Virginia ·
There is a particular kind of theater practiced in Sacramento, a blend of high-stakes urgency and polished vanity that Governor Gavin Newsom has mastered.
There is a particular kind of theater practiced in Sacramento, a blend of high-stakes urgency and polished vanity that Governor Gavin Newsom has mastered. His latest production, as reported by the Los Angeles Times, involves a "kill switch"—a phrase that sounds more like a plot point from a mediocre summer blockbuster than a piece of sound public policy. By issuing an executive order to rein in artificial intelligence "before it’s too late," Newsom isn't merely managing a technology; he is auditioning for a larger stage.
The Velocity of a Campaign Trail
The Governor’s rhetoric is breathless. He speaks of "urgent velocity" and "abject failure" in Washington, claiming that while President Donald Trump dismisses AI safety fears as a "hoax," California must build a "national model" to serve as the baseline for the rest of the country. As CNBC notes, this executive order demands a guide within two months to strengthen safety laws, including requiring independent third parties to write safety plans for "frontier" AI companies.
It is a convenient timing. Newsom’s term ends on January 4, and the whispers of a 2028 presidential run are no longer whispers; they are the primary melody. By positioning himself as the adult in the room—the one brave enough to pull the plug on a rogue machine—he creates a narrative of responsible leadership. He is following a well-worn path of executive overreach, echoing the US Executive Order on Safe, Secure, and Trustworthy AI from 2024, which sought to establish standards before the slow gears of Congress could even begin to turn. But Newsom isn't looking for a standard; he is looking for a spotlight.
The Moat Disguised as a Guardrail
The most capable advocates for this regulation point to the abyss. They cite researchers like Jacob Coxon, who warns that AI companies are "gambling with our lives," and the chilling reality that Anthropic’s Claude is already helping develop the next version of itself. To the sincere observer, the prospect of "humanity-ending capabilities" justifies any precaution. If the house is on fire, you don't argue about the cost of the water.
But look at who is cheering from the front row. The CEOs of the giants—Sam Altman of OpenAI and Dario Amodei of Anthropic—are not fighting these regulations; they are begging for them. This is the oldest trick in the corporate handbook: regulatory capture. By mandating expensive third-party audits and complex safety plans, Newsom is not protecting humanity; he is building a moat around the trillion-dollar incumbents.
Smaller startups, the very engines of American competition, cannot afford a priesthood of independent auditors to bless their code. We have seen this play before with the European Union General Data Protection Regulation and the more recent AI Act of the European Union, where complex compliance frameworks effectively subsidized the biggest players by pricing out the challengers. Newsom is simply importing the European model of stagnation and calling it "safety."
The Atomic Lesson of Centralized Control
The mechanism Newsom is proposing—transferring the "off switch" from the developers to a state-sanctioned regulatory body—is a ghost of the Regulation of the Atomic Energy Commission. After World War II, the U.S. moved the control of atomic energy from military hands to a civilian agency, the AEC, to foster peacetime development while maintaining a tight grip on the "switch."
The lesson of the AEC is not that centralization is safe, but that it creates a singular, political point of failure. When you move a high-risk, dual-use technology into a civilian regulatory body, you don't eliminate the risk; you simply change who gets to decide when the switch is flipped. In the 1940s, the stakes were national security. In 2026, the stakes are political leverage. By creating a framework where a panel of experts—appointed by the Governor—determines what constitutes a "rogue" model, Newsom is creating a political kill switch. He is establishing a precedent where the state can silence a technology not because it is dangerous, but because it is inconvenient or contrary to the prevailing orthodoxy of the day.
Governor Newsom is spending the tradition of American innovation to buy a few headlines and a perceived sense of control. He is trading the messy, vibrant competition of the garage startup for the sterile, managed safety of the corporate boardroom. In his rush to save us from a hypothetical machine, he is building a very real cage for the next generation of American builders.