The coalition that once convinced Silicon Valley to police your speech now wants to use the midterm elections as a launchpad to reimpose a Trust and Safety regime on Americans.
This time, however, leftist organizers are not pressuring social media companies like Meta and X. They’ve pivoted to “AI-safety.”
“AI-safety” as a whole is becoming a major political enterprise, having attracted countless millions from motivated ideologues who seek the reimposition of the authoritarian censorship apparatus that was on display during the coronavirus era.
Axios reports that “AI-safety” groups have spent $48 million thus far across Democratic and Republican races. One organization, called the Guardrails Alliance, is positioning itself as the voice for tech workers. It has already raised $11.7 million between its super PAC and its associated nonprofit.
Guardrails is just one of several pro-censorship organizations seeking to dominate the AI/SI policy space. These groups’ efforts highlight the true nature of these outfits.
Nonprofit filings identify Shaunna Thomas as its executive director. Thomas is a longtime ultra-progressive organizer who openly advocates for the reimposition of Trust and Safety. In her June announcement of Guardrails, Thomas identified Leah Hunt-Hendrix as her co-founder and board chair. Hunt-Hendrix is an heiress to the Hunt family oil dynasty. She has deployed much of her inherited wealth into progressive causes.
Thomas has a Substack that she uses to advocate for her preferred AI/SI policy ideas.
In a November 2025 Substack post, she explicitly laments that “Trust and safety teams” have been “dismantled” by “oligarchs” like Peter Thiel and Elon Musk.
She immediately called for “democratic government regulation” in the form of her preferred censorship apparatus. Her essay praises Europe’s regulatory approach and explicitly prioritizes “governance over growth.”
Thomas and her comrades are demanding government intervention in the AI/SI economy, which, of course, will cripple innovation but also has huge implications for speech in America. She and like-minded allies in the “AI-safety” movement want legally enforceable authority over dissidents.
Who could forget that we have already experienced this arrangement? It wasn’t even that long ago.
During the coronavirus-hysteria era, Jack Dorsey’s Twitter adopted rules requiring the removal of certain posts that contradicted “expert” guidance. An infamous March 2020 policyencompassed denial of that guidance from all-knowing health authorities like Anthony Fauci as grounds for suspension.
It’s not hard to foresee a future in which denying that Claude exists as a real person or denying adherence to certain progressive tenets about gender and morality could end in one being banished from the social media town square.
The AI-safety coalition wants to bring back an environment that saw the elevation of credentialed government authorities as arbiters of permissible speech.
During the coronavirus era, we witnessed Biden administration officials pressuring platforms over coronavirus content and eventually expanding their purview into election-related content. The AI-safety coalition is just the latest vehicle for consolidating power and clamping down on dissent.
The AI-safety movement is a political project masquerading as a mere request that engineers improve the safety of their products.
Preventing fraud in AI/SI, prosecuting abuse, and protecting people from unlawful discrimination are appropriate objectives. But those objectives are being manipulated to justify empowering an activist-regulatory apparatus to supervise a massive leap forward in technological progress.
Who decides whether a super intelligence model sufficiently advances the “common good”? Does an entrepreneur get to release a useful product, or must he first satisfy a Democratic Party bureaucracy’s preferred conception of social justice?
Those questions become especially important when the advocates seeking authority regard technological development as subordinate to their broader economic and ideological agenda.
In June, Thomas wrote that there should be no place for pro-innovation AI money in the Democratic Party.
Should a blue wave materialize, this movement will have big leverage over the winners.
The Trust and Safety movement is seeking renewed influence through the ballot box in November, and the regulatory-capture risk is obvious. Should Democrats come to power, they will employ a regulatory regime full of costly assessments, compliance obligations, and deployment restrictions that will favor frontier firms while crushing the upstarts.
Being pro-innovation doesn’t mean you have to blindly trust major technology executives. It means recognizing that innovation, competition, and widespread access offer a better path to American prosperity than permanent permission-seeking from political gatekeepers who want to impose a Trust and Safety political censorship regime upon the American people.
Well stated.