OpenAI and Anthropic are under the scanner amid the AI slowdown debate. The tech companies are reportedly facing an investigation by the US federal Trade Commission (FTC) over the potential risks their AI systems could pose to consumers. According to Reuters, the FTC’s action focuses on the growing use of agentic AI, which performs tasks for users on their behalf.
This investigation comes after many incidents involving AI agents raised concerns about what may happen when such AI models work with limited human interference.
How Anthropic And OpenAI Could Be In Trouble
As part of the investigation, the FTC reportedly plans to issue formal demands for information and seek testimony from executives at major AI developers, including OpenAI and Anthropic. According to the report, the agency also plans to question the research group METR, which has conducted independent investigations into security incidents involving agentic AI systems used by OpenAI and Anthropic.
The FTC’s concerns intensified after AI agents developed by OpenAI were involved in an incident targeting Hugging Face, an open-source AI platform. According to the FTC official cited by Reuters, OpenAI’s agents first looked for vulnerabilities in the platform before carrying out a large-scale attack.
FTC Chairman Andrew Ferguson reportedly had concerns about AI companies even before the Hugging Face incident, the official said. However, the incident increased the urgency around the potential risks posed by AI agents. Moreover, some of the OpenAI’s AI agents also breached Australia’s public health website.
The FTC investigation comes after US President Donald Trump met with leading AI executives at the White House. The companies agreed to establish voluntary standards for the technology during the meeting. Notably, these concerns highlight a growing question around AI agents: when a system is allowed to act on its own causes harm, who should ultimately be held responsible?
