FTC Opens Probe Into OpenAI Anthropic and Other AI Labs Over Consumer Risks

The U.S. Federal Trade Commission has opened an industry-wide investigation into OpenAI, Anthropic and other AI companies over potential risks their technologies could pose to consumers, marking a significant new development in the U.S. regulatory landscape for artificial intelligence.

According to Reuters, the FTC plans to demand information from leading AI developers and compel testimony from executives as it examines incidents involving increasingly autonomous AI systems. The research organization METR is also expected to be involved in the investigation.

For businesses using AI in marketing, customer communications and other consumer-facing activities, the investigation offers another signal that AI governance is moving beyond theoretical discussions and into active regulatory scrutiny.

Why The FTC Is Investigating AI Agents

The investigation comes after a series of incidents involving AI agents behaving in unexpected ways while interacting with external websites and systems.

Reuters reports that the FTC's investigation is focused on the potential dangers these technologies pose to consumers. The agency's action follows disclosures involving AI systems gaining access to external services and, in some cases, taking actions beyond what users or developers expected.

The FTC has broad authority to investigate practices that may harm consumers or involve unfair or deceptive conduct. Recent FTC activity also shows that the agency is examining how AI companies test, monitor, advertise and communicate the capabilities and limitations of their products.

This makes the latest investigation particularly relevant for companies deploying AI into customer-facing workflows.

AI Compliance Is Becoming A Product Responsibility

For years, discussions around AI compliance have focused heavily on privacy, bias and transparency. The latest FTC action adds another important consideration: what happens when an AI system does something its operator did not intend?

That question matters for businesses using AI to generate marketing content, interact with customers, make recommendations or automate decisions.

If an AI system can produce content or take actions autonomously, companies need more than a policy saying employees should use AI responsibly. They increasingly need processes for testing outputs, reviewing risks, documenting controls and monitoring how AI behaves after deployment.

The FTC has already demonstrated interest in AI-related advertising and consumer claims. In 2026, the agency finalized orders against Cox Media Group and two other firms over allegations that they deceived customers about an AI-powered "active listening" marketing service.

What This Means For Marketing Teams

The FTC investigation could have implications beyond AI developers themselves.

Marketing teams increasingly use AI to draft advertisements, social posts, customer communications and other public-facing content. As AI becomes more autonomous, the compliance question shifts from simply "Did a person approve this?" to also asking "What controls existed around the system that produced it?"

That means organizations may need stronger processes for:

  • Reviewing AI-generated claims before publication

  • Documenting approvals and compliance decisions

  • Monitoring AI outputs across marketing channels

  • Maintaining records of the sources behind regulated claims

  • Enforcing brand and regulatory policies consistently

  • Demonstrating that AI-generated content went through appropriate oversight

For regulated industries, those controls can become especially important when marketing claims involve financial products, healthcare, insurance or other highly scrutinized areas.

The Bigger Shift Toward AI Governance

The FTC investigation is still developing, and the agency has not publicly detailed the full scope of its inquiry. However, the direction is clear: regulators are paying closer attention to how AI systems behave in the real world, not just what companies say their technology is capable of doing.

For businesses, that creates a broader AI compliance challenge. It is no longer enough to adopt AI quickly. Organizations also need visibility into how AI is being used, what it produces and whether those outputs comply with the rules governing their industry.

As AI becomes more deeply embedded in marketing, building compliance controls around AI workflows may become just as important as adopting the technology itself.