Anthropic CEO Dario Amodei calls AI backlash a crisis of trust, not a messaging problem
Investor Gavin Baker told Anthropic CEO Dario Amodei in public to become a more positive industry advocate before the Anthropic IPO. Amodei refused on August 15-16 in an X thread that laid out a sharper diagnosis: the problem is not messaging, it is trust built over decades, and the cure is not a PR pivot.
What happened
Baker argued on the All-In podcast and on X that Amodei's safety rhetoric had fueled the AI backlash now slowing data-center permits and shaping regulatory proposals. Baker told Amodei he had "lost the argument" on regulation and, given that "he is about to be the CEO of one of the most important companies in the world," should "make an effort to be a more positive advocate for his own industry," per TechCrunch.
Amodei disputed that his writing has been skewed negative. His output was "about equally balanced between risks and benefits," he said, and his 2023 essay "Machines of Loving Grace" was written specifically because the AI industry was not "painting an inspiring enough picture of how the technology could radically transform the world for the better."
The core of his counter-argument: "I think it is fundamentally a crisis of trust. I think that ordinary people don't trust companies, governments, or the tech industry and always suspect that we are cooking up some new way to screw them over." The current AI backlash is, in his framing, the latest chapter of distrust built over decades rather than the product of any executive's public statements.
Amodei also redirected the sharpest critique inward. "By far the most accurate criticism of AI companies including Anthropic is that we haven't yet delivered on our big promises to benefit the world. That is totally on us." Promising to cure cancer is "more a cliche than it is inspiring," he wrote. Actually curing cancer would shift opinion in ways no communications campaign can.
On regulation, Amodei rejected what he called Baker's "false choice" between unregulated open distribution and regulatory capture by large companies. Anthropic's policy proposals, he said, aim to "disadvantage (slow down) frontier AI companies while advantaging smaller competitors." AI is "structurally a technology that tends to concentrate power," and open-weight models address this only partially: they "simply shift the concentration somewhat to those with the most compute and chip."
Why it matters
The trust framing has a concrete implication for operators choosing a model vendor. If public hostility is structural (decades of institutional skepticism, not tone-deaf executives), no messaging campaign repairs it. The credibility test for Anthropic and every frontier lab is product outcomes in medicine, productivity, and science, not X threads. That is a slower and harder bar than adjusting a CEO's public voice.
On open weights, the concentration argument cuts differently than the standard framing suggests. Compute-rich actors benefit most from open models, which weakens the "open equals decentralized" assumption many teams use to justify switching away from hosted APIs. The concentration just lands somewhere different.
Anthropic's stated design goal is rules that slow frontier labs while creating room for smaller competitors. Operators on proprietary APIs and those building on open-weight alternatives face different outcomes if those proposals influence legislation, and any policy text Anthropic publishes before the IPO filing will be the thing to read, not a social media exchange.
What to watch next
Watch for Anthropic's pre-IPO regulatory proposal text, specifically how it defines "frontier AI company" - that language will matter more than any X thread.