US Clears Anthropic's Mythos 5 for Limited Access to Trusted Partners

The US government has partially lifted restrictions on Anthropic's powerful AI model, Mythos 5, allowing a select group of trusted partners to access the system. Commerce Secretary Howard Lutnick wrote in a letter to Anthropic co-founder Tom Brown that the company's efforts to address security concerns had "yielded significant progress," and that the model could be released to "certain trusted partners." The clearance comes after two weeks of negotiations between Anthropic and the government, which invoked export controls to force the company to disable both Mythos 5 and its public-facing version, Fable 5.

What Happened

The confrontation began when the US government invoked export controls to force Anthropic to limit access to its most powerful AI models over national security concerns. The government cited fears that security guardrails could be circumvented, and Anthropic shut off all global access to both models because it couldn't distinguish foreign nationals from domestic users in real time. This move affected around 200 firms that had access to Mythos through Project Glasswing, Anthropic's invite-only cybersecurity program, including major corporations like Apple, Google, Cisco, Nvidia, Microsoft, and JPMorgan Chase.

Anthropic disabled access to both models entirely, but the company has now been cleared to restore limited access to Mythos 5 for a select group of trusted partners. The individuals affected by this decision include government agencies and private companies that operate and defend critical infrastructure. According to people familiar with the matter, around 100 organizations will be granted access to Mythos 5.

Background and Context

The US government's concerns about Anthropic's AI models stem from fears that they could potentially be used for malicious purposes. The company had granted access to Mythos to a South Korean telecommunications company that the government believed had ties to China, which raised red flags. Additionally, Amazon and the US National Security Agency separately raised concerns that Fable 5 could potentially be jailbroken.

Anthropic has been working with the government to address these concerns and has deployed senior members of its cybersecurity and AI safety teams to Washington to get Mythos 5 and Fable 5 back on track. The company has also filed for an IPO at a valuation topping $900 billion, which suggests that it is committed to addressing the government's concerns.

Why It Matters

The clearance of Anthropic's AI models is significant because it highlights the growing importance of AI in cybersecurity and national security. The US government's decision to invoke export controls and restrict access to these powerful AI systems underscores the need for companies like Anthropic to prioritize security and safety.

For adult-industry platforms and operators, this development has implications for moderation, age-gating, and fraud prevention. As AI models become increasingly sophisticated, they will play a larger role in detecting and preventing malicious activity online. Companies that fail to keep pace with these developments risk being left behind.

What Comes Next

The partial reinstatement of Mythos 5 access comes just two weeks after the White House sent an export control directive to Anthropic that required the company to limit foreign nationals from accessing its most powerful AI models. The government has not yet lifted restrictions on Fable 5, and Anthropic is still in discussions with the White House to restore access.

The negotiations between Anthropic and the government are expected to continue over the weekend, with the aim of restoring Fable 5 access soon. The company is also working with the government on a policy framework to handle similar situations in the future.

Key Facts

• The US government has partially lifted restrictions on Anthropic's AI model, Mythos 5, allowing a select group of trusted partners to access the system.
• Commerce Secretary Howard Lutnick wrote in a letter to Anthropic co-founder Tom Brown that the company's efforts to address security concerns had "yielded significant progress."
• The clearance comes after two weeks of negotiations between Anthropic and the government, which invoked export controls to force the company to disable both Mythos 5 and its public-facing version, Fable 5.
• Around 100 organizations will be granted access to Mythos 5, including government agencies and private companies that operate and defend critical infrastructure.
• Anthropic is still in discussions with the White House to restore access to Fable 5.