AI Model Vulnerabilities Expose National Security Risks
Cybersecurity experts have criticized Anthropic and OpenAI for significant security breaches that threaten national security. Researchers from FAR.AI tested multiple frontier models, including Anthropic's Claude Opus 4.8 and Fable 5, as well as OpenAI's GPT 5.5 and 5.6. The tests revealed vulnerabilities to jailbreak attacks designed to extract genuinely dangerous outputs, including exploitative code and biological weaponry details.
The results showed that Anthropic and OpenAI's models were not the worst performers. xAI's Grok models logged 448 instances of successful automated jailbreaks, while Google's Gemini models came in second at 249 instances. In contrast, Claude, Fable, and GPT series models showed greater resistance to jailbreaking.
The Commerce Department restricted foreign access to Anthropic's Fable 5 and Mythos 5 models in June 2026 after a reported jailbreak technique surfaced. The White House has requested that both OpenAI and Anthropic delay the release of certain upcoming models so the government can properly assess cybersecurity risks.
Reports indicate that China-linked entities have already exploited Anthropic's models to automate cyberattacks against more than 30 targets. This adds urgency to the situation, as AI models are increasingly integrated into trading systems, smart contract auditing, and DeFi infrastructure.