Anthropic CEO Responds to AI Backlash: At Its Core, It’s a “Crisis of Trust”

Anthropic CEO Dario Amodei has pushed back against claims that prominent AI leaders are fueling public backlash by speaking too often about the technology’s dangers. The criticism came after investor Gavin Baker argued that Amodei’s repeated warnings about AI risks may have contributed to growing resistance in the United States, particularly around the construction of data centers, and suggested that he should use his position to advocate more forcefully for the industry.

In a series of public responses, Amodei rejected the idea that his comments are the main driver of anti-AI sentiment. He said his approach has always been to discuss both the promise and the risks of AI, rather than presenting only one side. He pointed to his essay Machines of Loving Grace as an example of that effort, saying he wrote it in part because the broader industry has not done enough to communicate a compelling vision of how AI could improve people’s lives.

At the same time, Amodei acknowledged that negative public sentiment around AI is real and serious. But in his view, the problem runs deeper than whether executives strike the right tone in interviews or public statements.

The real issue, he argues, is trust

According to Amodei, the current resistance to AI is best understood as part of a much broader crisis of confidence. Public skepticism, he suggested, is rooted in longstanding distrust toward large corporations, government institutions, and the technology sector itself. From that perspective, the backlash surrounding AI is not a standalone reaction caused by a few cautionary remarks from industry leaders. Instead, it reflects a deeper and older deficit of trust that AI has now brought into sharper focus.

That framing shifts the debate away from messaging alone. If the public already doubts the motives and accountability of powerful institutions, then simply offering more optimistic talking points is unlikely to change much.

Promises are no longer enough

Amodei said the most valid criticism of Anthropic—and of the AI industry more broadly—is not that companies are talking too much about risk, but that they have yet to convincingly deliver on their sweeping promises about making the world better. In his telling, responsibility for that gap lies squarely with the companies themselves.

He argued that the industry should spend less time worrying about whether its public relations language sounds positive enough and more time confronting the substance of public criticism. Grand claims such as AI one day curing cancer, he suggested, have become familiar rhetoric. What will matter more is whether companies can produce real, verifiable outcomes that improve life in tangible ways.

For Amodei, trust will not be rebuilt through slogans or aspirational messaging, but through results people can actually see and evaluate.

Avoiding simplistic arguments on regulation

Amodei also addressed the increasingly polarized debate over AI regulation. He rejected the idea that the conversation must be reduced to two extremes: either AI should remain largely unregulated in order to spread widely, or any regulation will inevitably entrench the biggest players and concentrate power.

In his view, AI already has structural tendencies that can lead to concentration of power, which is precisely why carefully designed rules are necessary. He argued that an effective regulatory framework should not be seen as a barrier to progress by default. Instead, it can serve several purposes at once:

• help manage AI safety and alignment risks,

• place institutional limits on the dominance of frontier AI companies,

• and preserve room for open-source models to continue developing.

This position reflects a more nuanced approach than the common regulatory talking points often heard in the sector. Rather than treating oversight as inherently anti-innovation or assuming that all regulation benefits incumbents, Amodei is arguing for rules that address risk while also preventing excessive concentration.

What he wants the public debate to recognize

The broader message behind Amodei’s response is that the backlash against AI cannot be explained simply by saying executives scared the public. In his account, the resistance is emerging in a society where trust in major institutions was already weak. AI has become the latest arena where that distrust is being expressed.

If that diagnosis is correct, then the solution is not just better rhetoric from tech leaders. The industry will need to demonstrate credibility through action—by building systems that are safer, by showing measurable public benefit, and by proving that its ambitions are matched by accountability.

In other words, the path out of today’s AI skepticism is not likely to come from more polished optimism. It will come, if at all, from evidence that the technology industry can deliver on its claims in ways that people find believable.