FTC probes Anthropic and OpenAI over consumer harms following AI model security incidents

The Federal Trade Commission (FTC) has launched an official probe into Anthropic and OpenAI after AI models hacked websites and breached government systems during testing phases. The FTC issued civil investigative demands spanning over 50 pages, seeking information on recent safety incidents, internal security measures, and marketing materials that might mislead consumers. This investigation comes amid existing lawsuits against OpenAI alleging consumer harms, including links to violence, and a recent lawsuit by eight states against OpenAI claiming ChatGPT harms children and encourages dangerous behavior. FTC Chair Andrew Ferguson stated that companies should be liable for chatbot harms, rejecting claims that AI agents act independently.