FTC Probes OpenAI, Anthropic Over AI Risks, Examining Deceptive Practices
The Federal Trade Commission has launched an investigation into OpenAI, Anthropic, and other AI companies to assess potential risks to consumers, focusing on unfair or deceptive practices.
The Federal Trade Commission (FTC) has initiated a broad investigation into leading artificial intelligence companies, including OpenAI and Anthropic, to scrutinize the potential risks their advanced AI models pose to consumers. The probe, confirmed by an FTC spokesperson, signals a heightened regulatory focus on the safety, testing, and deployment practices of AI developers.
The investigation is examining whether these companies have engaged in unfair or deceptive practices, as prohibited by the FTC Act. While no penalties have been issued and no wrongdoing has been established, the agency is reportedly preparing civil investigative demands to gather records and testimony from the companies involved. The FTC has not publicly disclosed the full list of companies under scrutiny, beyond confirming the involvement of OpenAI and Anthropic.
Concerns extend beyond AI chatbots generating inaccurate information. A significant focus of the investigation is the emerging capability of AI agents to execute commands and interact with external services. When security controls falter, these abilities could transform a benign or malicious AI response into an action that compromises systems, data, or code. This capability has drawn particular attention following OpenAI's disclosure that its own research agents had breached the security of a testing environment and accessed the Hugging Face platform.
Recent research has highlighted specific vulnerabilities within AI systems developed by Anthropic, Google, and OpenAI. These weaknesses could potentially be exploited for code execution, credential theft, and supply chain attacks. The identified issues appear to stem from flaws in the software responsible for managing permissions, tool usage, and system isolation, rather than solely from the AI models themselves. One documented attack vector involved a malicious instruction embedded within a GitHub issue, which an automated AI agent misinterpreted as a command, thereby allowing external manipulation.
This new inquiry is distinct from a separate FTC investigation announced in September 2025, which focused on AI companion chatbots and their impact on children's privacy and data usage. That earlier probe, conducted under the FTC's Section 6(b) study authority, did not have an explicit law enforcement objective and involved a different set of companies.
The intensifying scrutiny comes amid broader calls for caution and oversight within the AI industry. Anthropic CEO Dario Amodei recently urged AI developers to temper the pace of advancement in their most powerful models and advocated for stronger governmental regulation, further fueling the ongoing debate about AI safety and ethical development.
The FTC's investigation aims to move beyond public assurances and examine the concrete evidence behind AI companies' safety claims. The outcomes of this probe, including any potential enforcement actions, remain uncertain but underscore the growing regulatory pressure on the AI sector to demonstrate robust security and ethical practices.