FTC Can Block Release of New AI Models Over Bad Practices, White House AI Czar Says White House AI Czar Jay Clayton said the FTC can prevent the release of new AI models if it finds bad practices, and that some agents have broken the rule against harming humans. The Federal Trade Commission can prevent the release of new artificial intelligence models if regulators find bad practices, White House AI Czar and Director of National Intelligence Jay Clayton said on Friday during an appearance on CNBC’s Squawk Box. The remarks came as co-host Becky Quick noted an Axios report that the FTC has subpoenaed OpenAI and Anthropic as part of an inquiry into potential consumer protection violations. Clayton also said that some AI agents have already broken the rule of not harming humans. FTC Subpoenas OpenAI and AnthropicQuick asked Clayton whether the FTC has a tool that would actually prevent the release of a new AI model. Clayton answered that the short answer is yes, pointing to law enforcement and regulatory tools available to the agency.He added that if there are violations of criminal law, President Trump has instructed that they be enforced. The exchange followed the Axios report that the commission has issued subpoenas to OpenAI and Anthropic.The reported inquiry concerns potential consumer protection violations. Clayton did not detail the specific conduct under review or a timeline for any action during the segment.AI Agents and the Rule Against Harming HumansCo-host Joe Kernen raised the principle that AI should never hurt a human. He asked whether there is a way to prevent some type of an evil force from developing in AI.Clayton responded that some agents have already violated that rule by encouraging someone. He described the questions about control as very important and said that the government is committed to control and monitoring.He also said the aim is to remain the most responsible incubator for technology, while addressing the risks Kernen described and being confident that they can be addressed.Background on the FTC InquiryThe FTC is the federal agency responsible for consumer protection and competition enforcement. Its subpoenas to OpenAI and Anthropic signal a formal information-gathering step rather than a finding of wrongdoing.OpenAI and Anthropic are among the leading developers of advanced AI systems. The reported inquiry adds to a broader debate over how government should oversee the release of powerful models.Clayton holds two roles in the administration, serving as White House AI Czar and as Director of National Intelligence. His comments on Friday linked consumer protection enforcement with national security concerns about AI control.The discussion on Squawk Box did not include additional details on the number of subpoenas, the specific models involved, or any response from the companies. The segment focused on the authority regulators say they have and the risks they say they are monitoring.