US govt clears limited access to Anthropic Mythos 5 AI model
New Delhi, June 27
US-based AI company Anthropic has won US approval to restore limited access to its powerful Mythos 5 artificial intelligence model to "certain trusted partners" after resolving the government's national security concerns.
Commerce Secretary Howard Lutnick wrote to Anthropic's chief compute officer that the company's efforts with the US government to address risks associated with the models "have yielded significant progress", multiple reports said.
The clearance follows a government order two weeks earlier that abruptly barred Anthropic from giving foreign nationals access to Mythos 5 and a related model, Fable 5, over fears that security guardrails could be circumvented.
The company responded by disabling global access and undertook intensive talks for a resolution.
"In just two weeks, we have worked diligently to ensure America remains the global leader in AI while safeguarding our security," Benno Kass, Commerce Department spokesman, was quoted as saying.
However, the recent letter doesn't mention any change to the government's restrictions on use of the Fable 5 model.
"We received notice from the US government that Mythos 5, our strongest cybersecurity model, can be redeployed to a small group of cyber defenders and infrastructure providers," Anthropic said in a statement.
The company said it is working to provision the approved set of providers and restore their access to Mythos 5 as quickly as possible. The company expressed optimism about the progress and said it will work with the government to expand access to Mythos 5 and make Fable 5 available for general use again.
Fable 5 is prevented from responding to some queries, such as to cybersecurity and biology related questions. In those cases, Anthropic has said its Claude chatbot will route responses through a different model, called Opus 4.8.
Anthropic had objected to the government's decision to impose export controls in a blog post, saying such standards applied across the industry, "would essentially halt all new model deployments for all frontier model providers."
— IANS
Reader Comments
I understand security concerns, but if Anthropic's models are truly safer than competitors, why limit access to only US entities? India has some of the best cybersecurity talent in the world - we could actually help strengthen these systems. Feels like another tech colonialism move, just like when they restricted chip exports to Huawei. At least they're partially restoring access now, albeit to a very small group.
Honestly, this is a double-edged sword. On one hand, I appreciate that they're being cautious about AI safety - we've seen too many sci-fi scenarios where unchecked AI goes wrong. On the other hand, the gatekeeping is obvious. The US wants to maintain its lead in frontier AI, and this "national security" excuse is convenient. Meanwhile, China is probably laughing and building their own version without any restrictions.
Interesting that Fable 5 is still restricted - the one that can't answer cybersecurity or biology questions. So the US government is basically saying "you can have the cybersecurity model, but not the one that might help develop better vaccines or defenses." The geopolitics of AI is becoming like the Cold War nuclear race, but with algorithms. India should invest heavily in our own foundational models - we have the IIT talent!
I work in cybersecurity in Bangalore, and this is frustrating. Mythos 5 could help Indian companies defend against ransomware attacks that are becoming more frequent. Instead, we'll have to rely on workarounds or older models. The US talks about "trusted partners" but then restricts access to even its allies. At least they're partially walking back their initial decision - the total ban was ridiculous.
N We welcome thoughtful discussions from our readers. Please keep comments respectful and on-topic.