In a dramatic turn of events, Anthropic is scrambling to lift a government-imposed ban on its cutting-edge AI models, Mythos 5 and Fable 5. The restriction, enacted due to alleged national security risks, was reportedly influenced by a conversation between Amazon’s CEO and senior officials in the Trump administration. This unexpected move has led Anthropic to disable its models for all users, including its own foreign employees.
A Sudden Halt for Advanced AI
Last week, the US government mandated that Anthropic cease the use of its Mythos 5 and Fable 5 models by foreign nationals, a decision that has sent shockwaves throughout the tech community. The firm was forced to act quickly, disabling its popular AI models in order to comply with the government’s directive.
According to Anthropic, this drastic measure was prompted by concerns over a potential method for ‘jailbreaking’ the Fable 5 model. In a statement, the company expressed its disagreement with the government’s decision, asserting, “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.”
Amazon’s Role in the Controversy
The roots of this ban trace back to a discussion between Amazon CEO Andy Jassy and Treasury Secretary Scott Bessent, as reported by the Wall Street Journal. Amazon’s researchers had raised alarms about the jailbreak issue, which led to concerns that it could be exploited for catastrophic cyber attacks. This warning from Amazon appears to have played a pivotal role in the government’s decision to enforce the ban.
Tech entrepreneur David Sacks, who advises the Trump administration, weighed in on the situation, stating on social media that the warning regarding the jailbreak was issued by a “highly credible trusted partner” of both Anthropic and the government. He further noted that the administration had urged Anthropic’s CEO, Dario Amodei, to address the jailbreak issue or withdraw the model entirely. However, Amodei’s refusal has intensified scrutiny on the company’s commitment to safety.
Broader Implications for AI and National Security
A separate report from Semafor indicated that the White House’s decision to restrict foreign access to Mythos was also influenced by fears over Chinese access to the AI technology. While Anthropic has proactively barred access to its models from China, the firm insists that it has not received specific concerns from the US government regarding adversarial use of the Mythos model.
This incident not only highlights the delicate balance between advancing technology and national security but also raises questions about the responsibilities of tech firms in safeguarding their innovations from potential misuse.
Why it Matters
The implications of this ban extend far beyond Anthropic and its AI models. As governments become increasingly vigilant about national security in the face of rapid technological advancement, companies in the AI sector must navigate a complex landscape of compliance and innovation. This incident serves as a stark reminder that while AI has the potential to revolutionise industries, it also presents unique challenges that require collaboration between tech firms and government entities to ensure safety and security.