The Trump administration has banned a leading AI model from customers, citing national security concerns over a jailbreak. Anthropic, the company behind the AI model, claims the vulnerabilities are minor and can be found in other models. The ban highlights the US government’s lack of transparency and consistency in regulating AI.
Experts say the administration should involve itself in conversations about AI safety, but their actions suggest a light approach to regulation. Some states have passed laws requiring AI companies to report safety issues and protect whistleblowers. Anthropic has been at the center of tensions with the Pentagon over its AI systems’ guardrails for military use.
The government’s decision has raised concerns that stifle the industry in the US, pushing experts to call for a clearer process for making cybersecurity decisions. “The problem is not that the government exercised discretion; national security demands such latitude,” said Jessica Tillipman. The Trump administration and Anthropic are meeting to resolve the issue, but some worry it sets a broader precedent.
AI regulation remains a pressing concern in the US, with no transparent, consistent framework for regulating AI models. “Right now, you have an ad hoc, personalized, opaque, possibly lawless approach,” said Brad Carson.
Source: https://edition.cnn.com/2026/06/21/tech/anthropic-ai-regulation