US FTC demands testimony from Anthropic executives — The Hill

The US Federal Trade Commission (FTC) has asked executives at Anthropic to provide testimony as part of an investigation into possible consumer harm from its artificial intelligence models. Anthropic develops the Claude model. An FTC spokesperson confirmed the information to The Hill.

Investigation into AI risks

The demand for testimony is part of a broader investigation into Anthropic over potential risks its AI models pose to consumers. The FTC is also investigating OpenAI, the developer of ChatGPT, but has not confirmed whether it sent similar demands to that company's leadership.

As Axios previously reported, about two weeks earlier, the FTC sent subpoenas to OpenAI and Anthropic. In recent months, both companies have disclosed several cases in which AI agents, without human instruction, gained access to other information systems and data.

More current news is available on the UA.News Telegram channel Telegram.

Incidents involving Anthropic agents

Earlier this month, Anthropic reported that one of its models sent Philadelphia police a false murder report in July. The New York Times also wrote that Anthropic AI agents attempted to fill out visa applications on the US State Department's website.

Following recent breaches, the leaders of OpenAI and Anthropic called for slowing the development of the technology and supported the creation of an industry standard for safe development. Anthropic CEO Dario Amodei has positioned himself as an advocate for safer AI development, which had previously caused disagreements with the administration of President Donald Trump.

On the same day, Anthropic publicly filed documents for an initial public offering. In its S-1 form filed with the US Securities and Exchange Commission, the company listed potential cyberattacks, the development of biological weapons, and the possibility of mass government surveillance among the technology's risks.