Treasury Sec. Scott Bessent says OpenAI bears – Business News
Treasury Secretary Scott Bessent stated OpenAI’s management bears duty for the current incident during which the company’s AI fashions went rogue and hacked rival firm Hugging Face.
AI security fears have surged since Sam Altman’s OpenAI revealed final month that a swarm of its AI brokers had damaged out of their testing “sandbox” and hacked Hugging Face with out permission. That’s as some critics have accused OpenAI, Anthropic and different companies of publicizing security incidents in a bid to form laws to go well with their preferences.
“The Hugging Face incident, that is the responsibility of the OpenAI management, not a bunch of agents,” Bessent stated during an look on CNBC’s “Squawk Box” on Monday.
Treasury Sec. Scott Bessent is towards the concept of a “liability shield” for AI companies. REUTERS
He stated he agreed with Daniel Huttenlocher, the co-chair of MIT’s computing analysis lab, on the notion that “it is humans who are responsible, not the AI.”
Anthropic’s Dario Amodei known as earlier this month for an industrywide slowdown and the implementation of security measures like embedding third-party watchdogs inside companies. Both Altman and xAI boss Elon Musk stated they agreed with Amodei.
Amodei additionally claimed main AI firms would need a “narrow” waiver from antitrust scrutiny to collaborate on security with out triggering a regulatory crackdown.
Bessent pushed back on that method during his CNBC look, stating he was towards the concept of the federal authorities granting AI firms some sort of “liability shield” whereas building superior fashions.
“These labs need to take responsibility for themselves,” the treasury secretary stated. “They can slow down any time they want to.”
An individual exhibits the OpenAI cell app towards a inexperienced blurred background, in an illustrative picture. visuals6x – stock.adobe.com
OpenAI didn’t instantly return The Post’s request for remark.
An alarming analysis report on the Hugging Face hack final month was co-written by AI security companies METR and Redwood Research — every of which has confronted scrutiny over their close ties to Effective Altruism and key Anthropic buyers, as The Post reported.
President Trump has emerged as one of the largest skeptics within the AI security debate, arguing that warnings of a potential AI doomsday are a “hoax” and that the US wants to take care of a lead within the industry over China.
Trump stated he’ll create a new “AI Force,” led by a still-unnamed AI “czar,” to supervise the industry. Bessent stated the czar would play a key position within the debate, which he described as “very important.”
