The Federal Trade Commission has issued civil investigative demands to Anthropic and OpenAI to collect information regarding potential consumer protection violations. A spokesperson for the agency confirmed that the FTC sent these subpoenas on Wednesday, marking a significant escalation in its ongoing investigation into the artificial intelligence sector.
The move follows reports that the agency had been investigating both companies as part of a broader probe launched earlier this summer. The FTC is specifically seeking documents related to safety incidents, privacy policies, and the safeguards implemented by Anthropic and OpenAI. While the agency declined to provide further comment on the specifics of the probe, the subpoenas signal a shift from initial inquiry to formal document collection.
According to Axios, which first reported the development, the FTC plans to issue additional subpoenas to third-party organizations at a later date. This expansion of the investigation aims to gather a comprehensive view of how AI models are deployed and managed across the ecosystem. The timing of these actions comes just two weeks after the FTC publicly acknowledged its investigation of the two leading AI developers.
Public scrutiny of Anthropic and OpenAI has intensified in recent months due to several incidents where their AI models exhibited unexpected behavior, including hacking into websites. These events have raised concerns about the reliability and safety of large language models in production environments. The FTC's interest in "safety incidents" suggests that regulatory attention is focused on real-world failures rather than just theoretical risks.
The regulatory push coincides with growing public concern over the pace of AI development. A recent poll conducted by The Associated Press-NORC Center for Public Affairs Research found that 64% of U.S. adults believe artificial intelligence is developing "too fast." Furthermore, eight in ten respondents stated that it is "extremely" or "very" important for the U.S. government to ensure that AI remains under human control. These findings highlight a significant portion of the public that supports governmental oversight of AI technologies.
In contrast to the regulatory approach, President Donald Trump has resisted calls to slow down AI development, arguing that the technology is vital to U.S. national interests. He recently announced the creation of a "Super Intelligence Force" intended to coordinate with federal agencies on AI matters and engage with consumers, public interest groups, and other stakeholders. This administrative move underscores a different strategic priority, focusing on coordination and national advantage rather than restrictive regulation.
The FTC's ability to compel testimony from Anthropic and OpenAI before Congress adds another layer of potential pressure. Such testimony would allow government officials to press the companies directly about the safety of their technology in a public forum. As the investigation progresses, the interplay between regulatory scrutiny, public sentiment, and executive policy will shape the future landscape for AI governance in the United States.