**
In a surprising turn of events, leading AI company Anthropic has temporarily withdrawn access to its latest models, Fable 5 and Mythos 5, following a directive from the US government aimed at enhancing national security. The company described the situation as a “misunderstanding” and is working diligently to restore access to its cutting-edge technology.
Sudden Withdrawal of AI Models
On Friday, Anthropic announced that it had to “abruptly disable” its advanced AI models for all users. This drastic measure was taken in compliance with an export control order issued by the US government, citing national security concerns. The specifics behind the directive remain murky, but Anthropic has indicated that the government believes there is a potential method to bypass safeguards intended to prevent Fable 5 from being misused to identify software vulnerabilities.
Interestingly, this development comes at a time when tensions between Anthropic and officials from the Trump administration were beginning to ease. Earlier this year, the relationship had soured as Anthropic refused to allow the military to utilise its AI models for domestic surveillance or autonomous weapon systems, resulting in the company being placed on a supply chain blacklist.
Growing Tensions Between AI Developers and Regulators
The recent order marks a significant escalation in the US’s efforts to restrict foreign adversaries’ access to AI capabilities. While export controls have traditionally focused on the hardware and tools that underpin AI technology, this is one of the first instances where direct access to AI models themselves is being curtailed.
Anthropic has expressed its dissent regarding the government’s decision. The company noted that it had only received “verbal evidence” of a potential narrow jailbreak and contended that this finding should not warrant a recall of a commercially deployed model that serves millions of users. The firm emphasised, “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.”
A Call for Balanced Regulation
Adding a layer of irony to the situation, just days before the export control order, Anthropic had advocated for increased oversight of AI technologies, including the authority to restrict models deemed to pose unacceptable risks. Despite this, the company maintains that Friday’s government action does not align with the principles of fair and evidence-based regulation.
Kirsten Davies, the Pentagon’s chief information officer, took to social media platform X to underscore the necessity of prioritising national security, stating, “Some things are simply more important than revenue cycles, clickbait, and pre-IPO valuation. America First. Always.”
Earlier this week, Anthropic had proudly launched its Claude Fable 5 model, showcasing new capabilities while incorporating safeguards to prevent misuse in sensitive areas like cybersecurity. Nevertheless, experts warn that if the Mythos models are misused, they could significantly boost the risk of sophisticated cyberattacks, particularly in industries that depend on outdated and complex technology systems.
The Path Forward for Anthropic
In light of the government’s directive, Anthropic is committed to resolving the situation as swiftly as possible. The company believes there is a fundamental “misunderstanding” regarding the nature of the concerns raised and is actively seeking to restore access to its models.
The abrupt halt affects only the Fable 5 and Mythos 5 models, with access to other Anthropic offerings remaining unaffected. The company has warned that if such stringent standards were applied industry-wide, it could result in a significant slowdown in the deployment of new AI models across the sector.
Moreover, Amazon’s cloud service, AWS, confirmed that Anthropic had requested the immediate revocation of access to its models for all users globally. A US official corroborated that the Commerce Department’s directive restricts access to Fable 5 and Mythos 5 for foreign nationals.
Dean Ball, a former White House official involved in the AI Action Plan, suggested that the order implies non-American users, even those residing in the US, may need to prove their citizenship to utilise Anthropic’s latest models. This raises questions about the potential implications for the company’s diverse talent pool, including co-founder Chris Olah and AI researcher Andrej Karpathy, both of whom hail from outside the United States.
Why it Matters
This episode highlights the growing friction between AI innovation and regulatory frameworks, underscoring the delicate balance that must be struck between national security and the free flow of technology. As governments grapple with the implications of advanced AI, the actions taken by Anthropic serve as a bellwether for the future of AI development and deployment in an increasingly contentious global landscape. The repercussions of this directive could shape the relationship between tech firms and regulators for years to come, impacting not just the industry’s trajectory but also the broader societal implications of AI technology.
