Anthropic has disabled access to two of its newest and most advanced AI models after receiving a directive from the U.S. government citing national security concerns.
Anthropic has taken its Fable 5 and Mythos 5 AI models offline after the U.S. government ordered the company to suspend access to the systems. The company said it received the directive and was instructed to block access to the two models because of concerns raised by government authorities. However, the order created a much bigger problem than a simple user restriction.
According to Anthropic, the directive applies to foreign nationals regardless of where they are located. It also affects foreign employees working within the company.
As a result, Anthropic said it could not practically separate affected users from the rest of its customer base and decided to disable both models completely while complying with the directive. “The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance,” the company said in a statement.
Anthropic stressed that the shutdown only affects Fable 5 and Mythos 5. Other models and services remain available. The decision comes shortly after the launch of Fable 5, which was introduced as one of Anthropic’s most capable public-facing AI systems. Mythos 5 was even more tightly controlled because of its ability to identify software vulnerabilities and security weaknesses.
Government officials have not publicly provided detailed information about the specific national security concerns behind the directive. However, Anthropic said it believes authorities were concerned about reports that users had found ways to bypass some of the safety protections built into Fable 5.
These techniques, often referred to as “jailbreaks,” are methods used to get AI systems to ignore certain restrictions or safeguards. Anthropic disputed the seriousness of the issue. The company said it reviewed the demonstration that appears to have triggered the government action and concluded that it involved only a limited number of previously known software vulnerabilities.
According to Anthropic, those vulnerabilities were not new discoveries and could already be identified by other publicly available tools. The company suggested that the concerns may have been overstated and described the situation as a misunderstanding. Despite that position, Anthropic said it had no choice but to comply with the directive while discussions continue.
The move has attracted attention across the technology industry because it shows governments are increasingly interested not only in controlling advanced computer chips but also in regulating access to the AI models themselves.
For years, much of the debate around AI regulation focused on the hardware needed to build powerful systems. Recent developments suggest regulators are now paying closer attention to what the software can do and who should be allowed to use it.
The shutdown of Fable 5 and Mythos 5 could also affect researchers, developers, and businesses that depend on advanced AI systems for their work. Many technology companies build products around these models, meaning a sudden suspension can disrupt projects, increase costs, and force teams to seek alternatives.
Industry analysts say the case could become an important example of how governments handle increasingly powerful AI systems. Supporters of tighter controls argue that advanced models could potentially be used to identify software weaknesses or assist cybercriminals if adequate safeguards are not in place.
Others worry that broad restrictions could slow innovation and limit access for legitimate users who rely on these tools for research and development. Anthropic remains one of the leading companies in the AI sector and competes directly with firms such as OpenAI and Google.
The company has invested heavily in developing advanced AI systems while also promoting safety measures designed to reduce misuse. This latest development places those efforts under renewed scrutiny.
For now, Fable 5 and Mythos 5 remain unavailable as Anthropic works through the implications of the government directive. The company says it hopes the situation can be resolved and that access to the models may eventually be restored.
Until then, users who relied on the two systems will have to wait while regulators and the company continue discussions behind the scenes. The episode highlights how quickly government decisions can affect access to some of the most advanced technologies currently available, even when the companies involved disagree with the reasoning behind those actions.
