The Federal Trade Commission launched a formal investigation into OpenAI, Anthropic, and other artificial intelligence companies, examining whether their product safety practices match public claims and regulatory obligations. The FTC issued orders to these firms demanding they submit documents and data related to how they test, monitor, and manage risks tied to their AI systems.
This regulatory action follows sustained pressure on leading AI developers after the Hugging Face security breach exposed vulnerabilities in the machine learning community's infrastructure. Hugging Face, a major platform hosting open-source AI models, disclosed unauthorized access to its systems in April 2023, raising questions about whether companies operating at scale have adequate safeguards in place.
The investigation targets OpenAI, the creator of ChatGPT and GPT-4, and Anthropic, the San Francisco startup behind Claude, among others. Both companies have positioned themselves as safety-conscious players in the AI space. OpenAI hired Sam Altman back as CEO after a brief board-led ouster in November 2023, partly to restore investor and public confidence in the company's governance. Anthropic has built its brand partly on claims of developing safer AI systems through constitutional AI methods.
The FTC's probe focuses on potential deception regarding AI capabilities, risks of misuse, and whether companies adequately protect against bias, privacy violations, and other harms. Regulators want to understand what safety testing occurs before models reach consumers and what happens after deployment when real-world problems emerge.
This investigation represents a shift from earlier FTC activity. Rather than wait for consumer complaints or documented harms, the agency is being proactive under Chair Lina Khan's leadership, examining business practices before serious damage occurs. The orders require companies to disclose competitive details about their safety infrastructure, which some industry participants view as intrusive but necessary for consumer protection.
The timing matters. Congress is debating AI regulation bills, and the Biden administration issued an executive order on AI governance in October 2023 calling for federal agencies to develop safety standards. The FTC's investigation provides policymakers with firsthand knowledge of how AI companies currently operate and where gaps exist.
Anthropic and OpenAI both said they welcome the inquiry and will cooperate fully. Both firms have boards with safety-focused directors and have published research on AI alignment and interpretability. However, public statements about commitment to safety and actual operational practices sometimes diverge, which is precisely what regulators want to verify.
The broader AI industry watches closely. If the FTC finds systematic safety failures, it could impose penalties, require operational changes, or refer cases to state attorneys general. More immediately, the probe signals that AI companies cannot rely solely on self-regulation and good intentions. Third-party audits, external safety testing, and transparent risk disclosure are becoming table stakes for operating in the AI market.
The investigation does not prevent these companies from launching new products or raising capital, but persistent regulatory scrutiny could affect investor sentiment, hiring, and partnerships. Investors in AI infrastructure and applications should track FTC filings and any enforcement actions that emerge.