📷 Image: Times of India (via NewsData)
Technology
US Govt Warns AI Firm Anthropic on Security Flaws in Advanced Models
✍️ Times of India
🗓 18 Jun 2026, 08:02 PM
👁 23
The U.S. government has instructed AI company Anthropic to address security vulnerabilities in its advanced models, threatening continued restrictions on releasing them internationally.
The U.S. government has presented artificial intelligence company Anthropic with a clear directive regarding the release of its advanced AI models, such as Fable 5, outside the United States. Officials are demanding that the company rectify identified security flaws before any international deployment is permitted.
This ultimatum follows an investigation by the National Security Agency (NSA) which confirmed vulnerabilities allowing for 'jailbreaking' of the AI models. The government expressed concerns that these exploitable loopholes could potentially be weaponized if not adequately addressed.
The administration's stance emphasizes the need for proactive patching of these security issues. However, some experts in the field have raised questions about the complete feasibility of eliminating all such vulnerabilities in complex AI systems.
This ultimatum follows an investigation by the National Security Agency (NSA) which confirmed vulnerabilities allowing for 'jailbreaking' of the AI models. The government expressed concerns that these exploitable loopholes could potentially be weaponized if not adequately addressed.
The administration's stance emphasizes the need for proactive patching of these security issues. However, some experts in the field have raised questions about the complete feasibility of eliminating all such vulnerabilities in complex AI systems.