AI companies have not delivered on big promises, says Anthropic CEO
Follow UsAnthropic CEO admits AI companies have not met many big promises.
He says real results are needed to build trust in AI.
Anthropic supports stronger rules for major AI companies.
Anthropic CEO Dario Amodei has admitted that AI companies, including his own, have failed to deliver on the ambitious promises they made about transforming the world for the better. Responding to investor Gavin Baker over a post on social media platform X, Amodei said this is the fairest criticism that can be levelled at the industry and one he accepts applies to Anthropic, too. He made the comment while defending his company against separate accusations that its regulatory positions concentrate power and that its public messaging on AI is overly negative. Amodei argued that real progress, not marketing, is what will restore public trust in artificial intelligence. He said Anthropic is speeding up work in biology and medicine and expects meaningful results in the years ahead.
SurveyThe exchange began after Baker, discussing Anthropic on a podcast, said he had been told Amodei privately suggested Anthropic could end up being the only private company left in the world, with only governments remaining alongside it. Fellow investor David Sacks called the remark hubristic, and Baker said he would discourage Amodei from repeating it.
Baker also argued that Amodei’s warnings about AI’s dangers were consistent with a broader case for concentrating the technology under a few regulated firms rather than distributing it widely and said this approach had failed to win the regulatory outcome Amodei wanted. Amodei took up the exchange on social media to respond to these points directly.
Also read: ChatGPT gets Google Drive support, reservation search and other new features: Here is what is new
Amodei rejected the idea that AI rules must either give more power to a few big companies or make AI widely available to everyone. He said Anthropic supports rules that put stricter requirements on leading AI companies like itself, while making things easier for smaller companies. He pointed to California’s SB 53 and safety testing rules being discussed with US regulators as examples.
In the X comment directed to Baker, Amodei disputed claims that he focuses too heavily on AI’s dangers. He clarified that his public output has been evenly split between highlighting risks and benefits, citing his essay ‘Machines of Loving Grace’, which argues AI could help cure most diseases within five to ten years. Moreover, he added that the social media clips shared widely often make his comments seem more negative than they actually were.
1/2 Thanks Gavin for an especially thoughtful exchange. I don't usually spend much time on social media but I wanted to engage here because it really brings out the heart of an important conversation.
— Dario Amodei (@DarioAmodei) August 15, 2026
First, on regulation, I think that “either concentrate it in the hands of a… https://t.co/2W6vWJAE8Y
Aside from that, he addressed why the public remains sceptical of AI, saying it stems from a broader, long-standing distrust of companies and institutions rather than from warnings issued by AI leaders. He pointed out that a polished marketing campaign would not fix that problem.
Also read: Teen accused of killing mother and brother reportedly used ChatGPT before the murders
He mentioned that his father died from hepatitis C a few years before effective treatments became available. Because of this, Amodei feels a strong personal need to help AI achieve real medical breakthroughs. Lastly, he said Anthropic will publicise any genuine results loudly once they arrive but will not make hollow claims in the meantime.
