FTC Opens Investigation Into OpenAI and Anthropic Over AI Safety

The U.S. Federal Trade Commission has opened an investigation into OpenAI and Anthropic as regulators assess whether leading AI companies have adequately disclosed the potential risks of their technology.

The investigation will examine possible unfair or deceptive practices and potential consumer harm. According to reports, the FTC is preparing formal requests for information and could compel executives at major AI developers to provide testimony.

The investigation comes at an important moment for the AI industry. Generative AI is rapidly moving beyond chatbots and becoming integrated into business software, search, e-commerce, customer service and autonomous agents.

As these systems gain more capabilities, the question is increasingly not only what AI can do, but who is responsible when it gets something wrong.

Existing Consumer Laws Meet Advanced AI

The FTC has broad powers to investigate unfair and deceptive business practices. Rather than immediately creating an entirely separate regulatory system for AI, the current investigation could test how existing consumer-protection laws apply to increasingly capable models.

FTC Chairman Andrew Ferguson has previously argued against rushing into new AI-specific regulations before determining whether existing laws can address emerging problems.

The agency has already been examining the AI industry from other perspectives.

An earlier FTC study looked at partnerships between major cloud providers and AI developers, including Microsoft and OpenAI, Amazon and Anthropic, and Google and Anthropic. The resulting report raised questions around competition, access to computing resources, switching costs and the concentration of technical and commercial information.

The latest investigation moves the discussion more directly toward consumer safety and the behavior of advanced AI systems.

AI Is Becoming Part of Ecommerce Infrastructure

For e-commerce businesses, these regulatory developments may seem distant from the everyday work of managing a product catalog or online store.

That distance is becoming smaller.

AI systems are increasingly involved in product discovery, conversational search, recommendations, customer support and content generation. More advanced agents can browse websites, compare products and take actions on behalf of consumers.

This means retailers and brands are beginning to rely on systems whose outputs they don’t fully control.

An AI shopping assistant, for example, might misunderstand a technical specification, recommend an incompatible accessory or incorrectly describe a product feature. The more consumers rely on AI to narrow their choices, the more important the accuracy of both the model and the information it receives becomes.

Reliable AI Needs Reliable Product Information

This creates a relevant connection with Icecat’s ecosystem.

AI models can interpret and summarize information, but the quality of the underlying product data still matters. Accurate identifiers, specifications, compatibility information, descriptions and product relationships give AI-powered e-commerce systems a stronger factual foundation.

Structured product information cannot eliminate AI errors. But it can reduce ambiguity in the source material that systems use to understand and compare products.

That distinction matters more as e-commerce moves from AI-assisted search toward agentic commerce.

When AI only suggests a product, an incorrect answer may inconvenience the shopper. When an AI agent is eventually trusted to select products or complete transactions, mistakes can have more direct consequences.

For businesses implementing these technologies, AI governance therefore needs to include questions about data provenance, accuracy and how automated outputs are verified.

Trust Could Become a Competitive Requirement

The FTC investigation does not mean OpenAI or Anthropic have been found to have violated consumer-protection law. It is an investigation into whether their conduct warrants enforcement, and its outcome remains uncertain.

But the investigation illustrates how the conversation around AI is changing.

For several years, competition centered primarily on which company could build the most capable model. As those models become embedded in everyday commercial activity, reliability, transparency and accountability are becoming equally important.

E-commerce is likely to experience that transition directly. AI systems will increasingly influence which products consumers discover, how they compare products, and potentially what they purchase.

Brands and retailers cannot control every AI system interacting with their products. What they can control is the quality of the information those systems encounter.

As AI takes on more responsibility in commerce, trustworthy AI and trustworthy product data will increasingly need to develop together.