Anthropic disables Fable 5 and Mythos 5 following U.S. Government directive

Artificial intelligence company Anthropic has abruptly suspended access to its most advanced AI models, Fable 5 and Mythos 5, after receiving a U.S. government directive citing national security concerns. The move comes just days after the public launch of Fable 5 and marks one of the most significant government interventions in the deployment of frontier AI systems to date.

Sudden order forces immediate shutdown

Anthropic said it received an export control directive from the U.S. Commerce Department requiring the company to suspend access to Fable 5 and Mythos 5 for all foreign nationals. Because the company could not reliably enforce the restriction based solely on citizenship status, it chose to disable the models for all users worldwide.

According to reports, Anthropic was given little time to comply before new export controls took effect. The company described the shutdown as abrupt and said all other Anthropic models remain available.

While the government has not publicly disclosed detailed technical findings, Anthropic said its understanding is that officials are concerned about the possibility that Fable 5 could be manipulated to assist in identifying software vulnerabilities that could be exploited in cyberattacks.

The directive reportedly stems from concerns surrounding the broader Mythos model family, which Anthropic has previously described as having advanced cybersecurity capabilities. The company had limited access to Mythos 5 through its Project Glasswing program because of concerns that the model could identify critical software weaknesses at an unprecedented scale.

Anthropic pushes back

Anthropic has publicly disagreed with the government's assessment, arguing that the evidence presented does not justify such sweeping restrictions. The company says the alleged vulnerabilities are neither unique to Fable 5 nor more severe than those found in other leading AI models.

In a statement, Anthropic said it believes there may be a misunderstanding and that it is working with government officials to clarify concerns and restore access as quickly as possible. The company warned that applying similar standards across the AI industry could significantly slow future model deployments.

The episode highlights the growing tension between rapid AI innovation and national security concerns. As governments grapple with the risks posed by increasingly capable AI systems, companies may face greater scrutiny over who can access their most powerful technologies and how those systems are deployed worldwide.