Anthropic on the AI Trust Crisis: Why Society Is Losing Faith in Tech Companies

The trigger was a statement by investor Gavin Baker — in the All-In podcast and on X, he wrote that Amodei’s warnings about the dangers of AI largely fueled a wave of negativity toward the industry in the US, especially toward data centers. Baker also noted that Amodei “lost the argument” on AI regulation — Anthropic supported a number of regulatory initiatives, including the California transparency bill for large AI companies — and advised him to become a more positive advocate for his own industry.

Amodei disagreed with the claim that Anthropic’s messages were disproportionately negative. The company’s texts maintain a balance between the risks and benefits of the technology, he stated, and he wrote the essay Machines of Loving Grace precisely because, in his view, the industry does not sufficiently highlight how AI can change the world for the better.

At the same time, Amodei acknowledged that society is negatively disposed toward AI and called this a serious problem. He sees the cause not in warnings from industry leaders, but in a deeper and longer-standing crisis of trust.

“Ordinary people do not trust companies, governments, or the tech sector and always suspect that we are plotting another way to deceive them,” Amodei said.

The most accurate criticism of AI companies, including Anthropic, is that they have not yet fulfilled their bold promises to bring benefits to the world, Amodei believes. He called this the fault of the industry itself and added that promises to cure cancer with AI sound like a cliché — only actual results will change people’s real attitude toward the technology.

In response to Baker’s thesis on regulation, Amodei said that it creates a false dilemma — between the free dissemination of AI without rules and the concentration of the technology in the hands of a few companies due to regulation. In Silicon Valley, there is a common simplified logic that “regulation equals regulatory capture equals concentration of power,” he noted, but he considers it too primitive — outside this circle, regulation is often perceived as a tool that curbs corporate power in the interests of citizens.

Anthropic consciously takes a cautious approach to legislative initiatives and strives to propose measures that rather slow down large AI companies than give them an advantage over smaller competitors, Amodei pointed out. He explained this approach by the fact that AI is structurally prone to concentrating power, and open-weight models do not fully solve this problem — they merely shift the advantage to those who have more computing power and chips.