Artificial intelligence company Anthropic has temporarily disabled access to its newly launched AI models, Fable 5 and Mythos 5, following concerns raised by United States authorities regarding potential national security risks. The decision has drawn significant attention across the technology sector, as the company described the move as a necessary step to comply with government directives while further assessments are carried out.
According to the company, the suspension affects all users and was implemented after officials expressed concerns about foreign access to the advanced systems. Anthropic stated that regulators believe a possible method may exist to bypass certain safety controls, creating fears that the technology could be misused or accessed in ways not originally intended by developers.
Fable 5 and Mythos 5 were introduced as some of Anthropic’s most powerful artificial intelligence models, designed to compete with leading systems in the rapidly expanding AI industry. The company had previously promoted the technology as a major advancement in reasoning, content generation, and problem-solving capabilities, making the sudden shutdown a notable development for customers and industry observers.
Meanwhile, Anthropic acknowledged that authorities have not publicly detailed the exact nature of the alleged vulnerability. However, the company indicated that officials suspect the existence of a technique capable of bypassing built-in restrictions, commonly referred to as a jailbreak, which could allow users to operate the models outside intended safeguards.
The latest dispute also reflects broader tensions between Anthropic and the US administration over the regulation and deployment of advanced artificial intelligence systems. Both sides have previously disagreed on restrictions involving government use of AI technologies, with legal challenges emerging over policies related to national security and procurement standards.
As discussions continue, the suspension highlights growing concerns about balancing innovation with security in the fast-moving AI sector. Industry experts expect regulators and technology companies to face increasing pressure to establish stronger oversight mechanisms while ensuring that advanced AI tools remain safe, reliable, and resistant to misuse.