Anthropic's CEO Says AI Backlash Is a Trust Problem, Not a Messaging Problem

Key points

• Investor Gavin Baker publicly argued on 15 August 2026 that Amodei's safety warnings have helped fuel a US backlash against AI and data centres.
• Amodei rejected the idea that his messaging is too negative, saying his writing is "about equally balanced between risks and benefits."
• Amodei called the public's distrust of AI "fundamentally a crisis of trust" in companies and the tech industry, decades in the making.
• Amodei said the fairest criticism of AI companies, including his own Anthropic, is that they "haven't yet delivered" on promises to benefit the world.
• On regulation, Amodei argued that rules can protect ordinary people from corporate power, not just slow innovation.

Dario Amodei, the chief executive of AI safety company Anthropic, has a message for the people who say he is scaring the public: it is not his fault, and the problem goes much further back than the AI boom.

The flashpoint was investor Gavin Baker, who appeared on the All-In podcast and posted on X (formerly Twitter) arguing that Amodei's habit of warning about AI dangers had helped stoke a backlash in the United States, including hostility toward the data centres that power AI systems. Baker told Amodei, as a leader of "one of the most important companies in the world," to become "a more positive advocate for his own industry."

Amodei was having none of it.

What did Amodei actually say?

He denied that his warnings have been one-sided. He pointed to his essay "Machines of Loving Grace" as proof that he actively champions the good AI can do, writing it precisely because he felt the industry was not painting an inspiring enough picture of AI's potential benefits.

But on the core question, he was blunt. "I think it is fundamentally a crisis of trust," he wrote in his public reply. "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."

In his reading, this is not a new problem that AI created. It is a decades-long breakdown in public confidence, and the current AI backlash is "just the latest iteration of it."

Is he taking any responsibility?

Yes, on one specific point. Amodei said the sharpest criticism of AI companies, Anthropic included, is straightforward: "We haven't yet delivered on our big promises to benefit the world." He called that "totally on us" and said it is the criticism worth making, not arguments about PR and messaging.

First reported by TechCrunch, the exchange also touched on regulation. Baker suggested that rules on AI tend to concentrate power in a few big companies by raising barriers that only giants can clear. Amodei called that a "false choice" and pushed back on what he described as Silicon Valley shorthand where "regulation equals regulatory capture equals concentration of power."

"Many people outside this bubble," he wrote, "think of regulation as something that constrains corporate power and benefits ordinary people."

He added that Anthropic tries to design its own policy proposals to slow down large frontier AI companies, the biggest players at the cutting edge, while giving smaller competitors room to grow.

What does this mean for ordinary people?

If Amodei is right, no single press release or reassuring speech will fix public scepticism about AI. The companies building these systems need to show real, tangible benefits in people's lives before trust follows. Watch for what AI companies actually deliver over the next few years, not just what they promise.