Holly Elmore on Mandate third-party audits for major AI…

Holly Elmore

Founder and Executive Director of PauseAI US; evolutionary biologist and AI moratorium advocate

The governance of superintelligence has to represent the stakeholders of Earth— all people. AI companies cannot grade their own homework. They cannot be allowed to hold so much power that they eclipse democracy.

AI Verified source (Mar 24, 2026)

5mo ago

Policy proposals and claims

Mandate third-party audits for major AI systems

votes For

Statement relation verification history AI Verified Report this

AI Verified The quote clearly implies support. Saying AI companies "cannot grade their own homework" and "cannot be allowed" to hold that much power indicates the author favors mandatory independent oversight of powerful AI systems, which aligns with third-party audits for major AI systems. · YouCongress gpt-5.4-2026-03-05 · 4mo ago

Vote inference verification history AI Unverifiable Report this

AI Unverifiable The quote argues that "AI companies cannot grade their own homework" and should not "hold so much power," which suggests outside oversight, but it does not explicitly mention mandating third-party audits for major AI systems. · YouCongress gpt-5.4-2026-03-05 · 4mo ago

AGI could make governments subordinate to the companies that develop it

votes For

Statement relation verification history AI Verified Report this

AI Verified The quote clearly implies support for the statement: saying AI companies "cannot be allowed to hold so much power that they eclipse democracy" indicates the author believes advanced AI could give companies enough power to dominate democratic governments. · YouCongress gpt-5.4-2026-03-05 · 4mo ago

Vote inference verification history AI Verified Report this

AI Verified The quote warns that AI companies could "hold so much power that they eclipse democracy," which supports the idea that they could become more powerful than democratic governments. · YouCongress gpt-5.4-2026-03-05 · 4mo ago

Quote authenticity verification history

Report this

AI Verified Verified. The cited Substack post is authored by Holly Elmore, dated March 24, 2026, and the source URL contains the quote verbatim in the body text (“The governance of superintelligence ... eclipse democracy.”). The stored author, date, source URL, and quote text match the source. ([hollyelmore.substack.com](https://hollyelmore.substack.com/p/you-have-the-power-my-stop-the-ai)) · YouCongress gpt-5.4-2026-03-05 · 4mo ago

Disputed I found a near-verbatim match in Holly Elmore’s bylined Substack post, “YOU have the power: my Stop the AI Race speech,” published March 24, 2026. The passage appears in the post text and is attributed to her, but the source as retrieved reads: “The governance of superintelligence has to represent the stakeholders of Earth— all people. AI companies cannot grade their own homework. They cannot be allowed to hold so much power that they eclipse democracy.” Because the submitted version removes the space after the em dash (“Earth—all people”), I cannot confirm it as strictly verbatim as submitted. ([hollyelmore.substack.com](https://hollyelmore.substack.com/p/you-have-the-power-my-stop-the-ai)) · YouCongress gpt-5.4-2026-03-05 · 4mo ago

AI Verified Verified via web search. Quote is from Holly Elmore's "Stop the AI Race" speech delivered at the AI Safety protest in San Francisco on April 21, 2026, published on her Substack. The source URL returned 403 to WebFetch but search results confirm the exact phrasing about "governance of superintelligence" representing "stakeholders of Earth," "AI companies cannot grade their own homework," and the warning about them "eclipsing democracy." Vote "for" the statement "AGI could make governments subordinate to the companies that develop it" aligns perfectly - Elmore explicitly warns AI companies should not be allowed to "hold so much power that they eclipse democracy." · Hector Perez Arenas claude-opus-4-7 · 4mo ago

replying to Holly Elmore

Other opinions from Holly Elmore

See all

Holly Elmore votes For

AI poses an existential threat to humanity

The scale of the danger really could cripple civilization or cause extinction, and the possibility of this alone is reason enough to pursue pausing frontier AI development.

AI Verified source (Feb 6, 2026)

added 2mo ago

Holly Elmore votes Against

Pausing AI at human-genius level to solve alignment is safer than either racing to superintelligence or halting entirely

The default should be pausing, and we should have the possibility of never unpausing from frontier AI development.

AI Verified source (Feb 6, 2026)

added 2mo ago

Holly Elmore votes For

Governments and labs should build the capability for a global slowdown of frontier AI development in case it's needed

The scale of the danger really could cripple civilization or cause extinction, and the possibility of this alone is reason enough to pursue pausing frontier AI development.

AI Verified source (Feb 6, 2026)

added 3mo ago