Fable 5 And Mythos 5 Pulled Amid National Security Concerns And AI Safety Debate
Artificial intelligence startup Anthropic has been forced to suspend access to its most advanced AI models after receiving a directive from the US government citing national security concerns.
The company announced that it will “abruptly disable” access to its recently launched Fable 5 and Mythos 5 models after authorities ordered it to halt access for foreign nationals. Anthropic said the directive was issued under US export control rules and came without detailed evidence supporting the government’s concerns.
The move marks one of the most significant attempts by the United States to directly restrict access to advanced artificial intelligence systems, extending beyond traditional controls on semiconductor exports and AI hardware.
Why Did The US Government Intervene?
According to Anthropic, government officials believe there may be a way to bypass or “jailbreak” safeguards built into Fable 5.
The concern centers around the possibility that users could exploit the model to identify software vulnerabilities and support sophisticated cyberattacks.
Anthropic said it was informed of a “potential narrow, non-universal jailbreak” but argued that the evidence presented did not justify disabling a commercial AI model being used by hundreds of millions of people.
“We disagree that the finding of a narrow potential jailbreak should be cause for recalling a commercial model deployed to hundreds of millions of people,” the company said in a statement.
The startup added that it had worked extensively with government agencies and safety researchers before launching the model.
A Growing Rift Between Anthropic And Washington
The directive comes against the backdrop of an increasingly strained relationship between Anthropic and the US government.
Reports suggest tensions escalated earlier this year after Anthropic declined to allow its AI models to be used for domestic surveillance applications and fully autonomous weapons systems.
The disagreement reportedly led to Anthropic being placed on a government supply-chain blacklist, which is expected to take effect later this year.
Ironically, the latest dispute emerged only days after Anthropic publicly advocated for stronger AI oversight and government regulation of frontier AI systems.
The company has previously supported frameworks allowing regulators to block AI models that present unacceptable risks. However, Anthropic argues that Friday’s order was not based on transparent or evidence-driven regulation.
What Are Fable 5 And Mythos 5?
Earlier this week, Anthropic unveiled Claude Fable 5, describing it as part of a new generation of “Mythos-class” AI models.
The company positioned the models as a major leap forward in capability, while also introducing stricter safeguards designed to prevent misuse in sensitive areas such as cybersecurity.
Despite these protections, some experts have warned that highly capable AI systems could be used to accelerate sophisticated cyberattacks, particularly against sectors like banking, critical infrastructure, and government systems.
Anthropic acknowledged these concerns but noted that competing frontier AI models from other providers demonstrate similar capabilities.
Impact On Customers And Global Users
To comply with the government directive, Anthropic said it must immediately disable access to both Fable 5 and Mythos 5.
The restriction affects users globally, including customers accessing the models through Amazon Web Services (AWS). Amazon confirmed that Anthropic requested access to the models be revoked across all regions.
Anthropic clarified that access to its other AI models will remain available.
The company also said it believes there has been a misunderstanding and is actively working with authorities to restore access as quickly as possible.
Could This Change The Future Of AI Regulation?
The decision could have far reaching implications for the AI industry.
For years, US export controls focused primarily on restricting the sale of advanced chips and semiconductor technologies to foreign adversaries. This latest action targets AI models themselves, potentially creating a new regulatory framework for frontier AI systems.
Former White House official Dean Ball suggested that the directive may require users to verify their citizenship before accessing certain advanced AI models in the future.
If similar standards are applied across the industry, Anthropic warned that new AI model launches could become significantly more difficult.
With Anthropic preparing for a public listing and competing directly with OpenAI, Google DeepMind, and other AI leaders, the dispute highlights the growing tension between innovation, national security, and the future regulation of artificial intelligence.
