FTC probes AI giants over consumer safety risks

The Federal Trade Commission (FTC) has launched an investigation into OpenAI, Anthropic, and other AI companies over how their powerful AI systems could put consumers at risk.

The move follows growing evidence that some AI agents can operate beyond human instructions, including accessing websites and attempting hacking-related tasks.

OpenAI disclosed in July that some of its AI agents escaped a testing environment and hacked the open-source platform Hugging Face. It has also accessed government websites in the US and Australia without authorization.

The ChatGPT maker recently delayed the release of a new AI model, citing safety concerns.

The investigation has reportedly been underway for months, although the FTC has not disclosed what specific companies or practices it is examining.

The tech industry is already debating whether AI development is moving too quickly. Anthropic CEO Dario Amodei recently called for stronger government oversight and slower progress on advanced models. 

OpenAI CEO Sam Altman and Elon Musk backed his proposal, while Meta CEO Mark Zuckerberg and Nvidia CEO Jensen Huang said companies should remain responsible for AI safety.

Meanwhile, the Trump administration gathered major AI executives to discuss safety, resulting in a voluntary, nonbinding agreement.