Scott Bessent criticizes AI leaders’ warnings as alarmist

US Treasury Secretary Scott Bessent has a message for the AI executives warning that their own products could end civilization: fix it yourselves.

In an interview on October 3, 2026, Bessent called the industry’s doom-laden warnings alarmist. He argued that sounding the alarm without offering solutions does not amount to leadership.

The comments matter because they come from the official who oversees the US financial system. They also arrive while Washington and Beijing are discussing AI safety mechanisms.

What Bessent said

Bessent’s core complaint was aimed at prominent AI figures who talk about existential risk. In his view, they keep issuing warnings without putting forward workable answers.

To make his point, Bessent reached for a horror-movie comparison. He likened the control-risk debate to Hannibal Lecter, the fictional serial killer.

“Stop me before I kill again.”

His preferred answer is self-regulation. He urged AI companies to take proactive steps on their own rather than wait for lawmakers to impose restrictions.

That does not mean the government is stepping back entirely. Bessent said Washington should remain vigilant about potential AI threats.

He drew one firm line, though. Bessent confirmed the government will not give AI developers a “liability shield.”

His reasoning was simple. Accountability, he insisted, rests with humans, not with AI systems.

How Bessent got here

The October remarks are not a sudden pivot. They line up with Bessent’s earlier warnings against liability waivers, which he laid out in a series of statements in September 2026.

Also in September 2026, Bessent advocated for a voluntary pause in the development of frontier AI models.

Bessent has been raising AI policy concerns since spring 2026. Those concerns followed warnings about risks to critical infrastructure.

The administration has stressed guarding against advanced AI models that could undermine the financial system, while also wanting the US to keep its technological lead.

The backdrop is ongoing US-China talks on AI safety mechanisms.

What this means for AI companies and investors

The most direct stakeholders are AI developers, and the signal to them is mixed. On one hand, a Treasury Secretary favoring self-regulation over new restrictions can read as friendly to innovation.

On the other hand, the refusal to offer a liability shield cuts the other way. Without that protection, developers carry the legal and financial exposure if their systems fail.

Bessent’s call for a voluntary pause on frontier models is another open question. It is unclear how a voluntary pause squares with a national goal of staying ahead, especially with a rival like China in the frame.

The financial-system angle deserves attention from banks and market participants. Bessent has tied his AI concerns to the stability of that system, which suggests Treasury will keep examining how advanced models interact with finance.

Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.