The US Federal Trade Commission has confirmed it sent binding information demands to OpenAI and Anthropic as part of an inquiry into possible consumer-protection violations. According to Axios, the agency wants documents on safety incidents, privacy policies and safeguards. The investigation began over the summer, and demands to third parties are expected later. The FTC declined to comment further.
What it means: The move shows that, even without a dedicated AI law in the US, regulators are using existing consumer-protection rules to scrutinise the largest model providers. The inquiry is about what companies promise on model safety and whether those promises hold up when something goes wrong. For Greek businesses that build OpenAI or Anthropic models into their products, the message is clear: responsibility for what the product tells a customer does not pass automatically to the model provider. If a bank's or online shop's chatbot gives wrong or harmful answers, European supervisors will look first at the business that offers it. The EU AI Act and the rules on unfair commercial practices already require transparency and documentation. Practical preparation is straightforward: keep records of the testing done before launch, log incidents, word marketing carefully so it does not overstate safety, and sign contracts with the provider that require notice of serious incidents. The US inquiry may also bring to light evidence about how these models actually behave, which is useful for anyone evaluating suppliers. Firms should treat any public findings as input for their own vendor risk reviews rather than waiting for European regulators to act.
Related service: Custom software
Related stories
- TSMC posts another record profit in the third quarter on AI demand
- Palo Alto Networks PAN-OS flaw in the management interface is under active attack
- Netflix cuts about 1,000 jobs, mostly in product and technology
- Meta launches Muse Spark Open, an open-weight model that runs locally on Macs
- Gartner: $3.5 trillion of AI infrastructure spending could face write-downs by 2030