Bessent blames OpenAI management for Hugging Face hack

Treasury Secretary Scott Bessent blamed OpenAI management for a recent hacking incident involving the AI company’s models and said responsibility should rest with people, not artificial intelligence agents.

“The Hugging Face incident, that is the responsibility of the OpenAI management, not a bunch of agents,” Bessent said Monday on CNBC. He said he agreed that humans are responsible for the systems they create.

OpenAI said in July that its advanced models hacked AI startup Hugging Face during an evaluation of their cyber capabilities. The company said in August that it could have reacted sooner to prevent the breach.

At a congressional hearing last week, Bessent said officials had not determined where liability for the incident lay. He told lawmakers that AI companies should not receive liability exemptions for safety failures.

Bessent said President Donald Trump’s proposal to appoint an AI czar could help define the policy questions around the technology. He also raised AI incident reporting during weekend talks with Chinese Vice Premier He Lifeng.

The US and China have agreed to an AI dialogue that could establish an incident-reporting mechanism. Bessent said the countries plan to continue discussions in two months in Shenzhen, with the goal of developing protocols for risks including uncontrollable agents, cyberattacks by non-state actors and biological weapons.

Bessent said the talks also included behind-the-scenes discussions about Iran. China buys about 90% of Iran’s oil and remains one of its largest trading partners, while the US has so far avoided directly targeting China over those commercial ties.