US Government and Anthropic Fail to Reach Agreement: Claude Fable 5 Model Remains Restricted

Negotiations between the US administration and Anthropic, one of the leading companies in the field of AI, have ended without result. During a meeting in Washington, the parties failed to reach an agreement regarding the new Claude Fable 5 model. As a result, export restrictions imposed on this model due to cybersecurity concerns will remain in effect. This is reported by Ixbt.com news reports.

The negotiations were attended by representatives of the US Department of Commerce, the Office of the National Cyber Director, and the AI Safety Institute. From Anthropic, co-founder Tom Braun and lead security researcher Nicholas Carlini participated, but they were unable to convince government officials that the model is completely safe. The fact that Commerce Secretary Howard Lutnick joined the communication remotely indicates how serious the issue is.

Security Vulnerabilities and Cyber Risks

The main disagreement concerns the possibility of "jailbreaking" the Claude Fable 5 model, meaning the ability to bypass established security barriers through special prompts. According to US government analysis, there is a risk of unauthorized access to the hidden and expanded capabilities of the Mythos system upon which this model is based. This could be used for cyberattacks or generating hazardous information.

Audit results conducted by the US National Security Agency (NSA) indicated that the model's protection mechanisms are indeed weak. Anthropic representatives argue that these concerns are exaggerated and that the model's actual level of protection is much higher. A group of independent experts supported the company's position in an open letter to the government, but this did not influence the decision.

A New Precedent for the Industry

The situation surrounding Claude Fable 5 is expected to be a significant turning point for the AI industry. From now on, large laboratories will be forced to work more closely with state regulators before releasing their products to market. This could fundamentally change the balance between technological freedom and state control.

For now, the public release and export of the Claude Fable 5 model have been postponed indefinitely. For Anthropic, this is not only a technological but also a financial blow, as investors are viewing these new restrictions with concern against the backdrop of the company's ongoing disputes with the Pentagon over military projects. The model may only be returned to consumers once all security flaws are eliminated.