InfoSecBulletin Cybersecurity for mankind
Anthropic said on Friday it will quickly turn off its best AI models for everyone. This comes after the U.S. government told it to stop letting foreign nationals use the models, because of security worries.
Hackers Exploiting MikroTik RouterOS Flaw Gaining Network Access
CrowdStrike launches SafeMind, the first agentic cybersecurity solution for defenders
Claude, DeepSeek, and Qwen AI agents combinedly used to hit gov.t across Asia
India: C-DOT Launches 14 Local Quantum-Safe Technologies
14,000 Dahua cameras compromised: TP-Link flaws enable RCE
Microsoft Reveals Project Zenith Windows PCs Able to Run 30B+ AI Models Locally
Google issues warning of new Chrome zero-day flaw exploited
CrowdStrike’s ‘FalconFlank’ zero-day allows SYSTEM privileges
727,000 data exposes: French hospital fined €500,000
153 Million Driver’s License Surfaced on Dark Web: FBI Starts Investigation
The company received the export control directive to suspend access to Fable 5 and Mythos 5 for all foreign nationals, without being given specific details of its national security concern, Anthropic said in a statement.
It is Anthropic‘s understanding that the government believes there is a method of bypassing, or “jailbreaking,” a safeguard that would prevent Fable 5 from being used in identifying software vulnerabilities, the company said.
The action is a big step up in U.S. efforts to stop other countries from using AI. For years, U.S. rules have mainly targeted the chips and tools that drive AI instead of limiting access to AI itself.
Anthropic said the government has given it only “verbal evidence of a potential narrow, non-universal jailbreak”.
“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,” the company said.
The government order and Anthropic’s reply show increasing conflict between AI makers and regulators on how to evaluate risks from “jailbreaks,” or ways to get around model protections.
As of Wednesday, Anthropic asked for more U.S. control over AI. They wanted the power to stop models with big risks. But they said the government’s action on Friday did not follow fair and fact-based rules.
The Pentagon’s chief information officer, Kirsten Davies, said in a post on X that the Defense Department supported prioritizing national security.
“Some things are simply more important than revenue cycles, clickbait, and pre-IPO valuation. America First. Always,” Davies said.
Anthropic secretly applied for a U.S. IPO last month, moving ahead of competitor OpenAI to enter public markets.
Earlier this week, Anthropic launched an AI model called Claude Fable 5, which is a new level of ability called “Mythos-class.” This model has protections that stop it from being used in dangerous areas like cybersecurity. Some users say these protections are “overly broad,” according to Anthropic.
