What does the new security guardrail do?

When users try to access restricted capabilities related to cybersecurity and biology, they are notified their request is blocked and it is processed by the less-advanced Opus 4.8 AI model instead. The safeguard extends to requests related to a specific behavior identified in an Amazon paper, addressing a workaround where users could bypass restrictions by asking the model to fix code rather than identify security issues.

Is Anthropic's regulatory situation fully resolved?

No. While the Commerce Department lifted export controls on Fable 5 and Mythos 5, Defense Secretary Pete Hegseth has told advisers there is no clear path to lift his February 28 order designating Anthropic a supply chain risk.

Back to articles

Trump Administration Lifts Export Controls on Anthropic's Claude Models

WIRED AI13h ago5 min read

Key takeaway

The Trump administration lifted export controls on Anthropic's Claude Fable 5 and Mythos 5 AI models after the company agreed to add a security guardrail that blocks access to certain restricted capabilities and routes those requests to a less-advanced model. The guardrail extends an existing protection to requests related to a specific behavior identified in an Amazon paper that researchers had found users could exploit. However, Anthropic still faces a separate supply chain risk designation from the Defense Department with no clear path to removal.

Summaries like this, in your inbox every morning.

3 Key Points

What happened
The Trump administration removed export restrictions on Anthropic's Claude Fable 5 and Mythos 5 AI models after the company agreed to extend a security guardrail. The new safeguard blocks users from accessing certain restricted capabilities related to cybersecurity and biology, routing those requests to the less-advanced Opus 4.8 model instead.
Why it matters
The export controls had effectively taken Fable 5 offline after the administration learned that users could bypass restrictions by asking the model to fix code rather than identify security issues. Commerce Secretary Howard Lutnick's decision to restore access signals that the administration believes Anthropic's new safeguards are robust enough to address those concerns, though the company still faces a separate designation as a supply chain risk from the Defense Department.
What to watch
While Anthropic has resolved the Commerce Department impasse, Defense Secretary Pete Hegseth has indicated there is no clear path to lift his February 28 order designating the company a supply chain risk, meaning the company's regulatory challenges with the administration are not entirely resolved.

FAQ

What does the new security guardrail do?: When users try to access restricted capabilities related to cybersecurity and biology, they are notified their request is blocked and it is processed by the less-advanced Opus 4.8 AI model instead. The safeguard extends to requests related to a specific behavior identified in an Amazon paper, addressing a workaround where users could bypass restrictions by asking the model to fix code rather than identify security issues.
Is Anthropic's regulatory situation fully resolved?: No. While the Commerce Department lifted export controls on Fable 5 and Mythos 5, Defense Secretary Pete Hegseth has told advisers there is no clear path to lift his February 28 order designating Anthropic a supply chain risk.

Discussion

No discussion yet for this article

Stay ahead with AI news

Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.

Get Started Free

Free · takes 30 seconds · unsubscribe anytime

1 minute a day. The AI essentials.

200+ sources · Email / LINE / Slack

Get it free →