The U.S. Department of Commerce has lifted export restrictions on the Claude Fable 5 and Claude Mythos 5 models. The Anthropic team officially confirmed this news on June 30, announcing that Fable 5 will once again become globally available starting July 1.

The model will be integrated into Claude Platform, Claude.ai, Claude Code, and Claude Cowork. Access via AWS, Google Cloud, and Microsoft Foundry will be restored as soon as possible. This decision resulted from productive negotiations with U.S. authorities.

A New Safety Standard

To bring back Fable 5, developers trained a specialized safety classifier. Its task is to block actions described in Amazon's report to the government. If a request is rejected, the user receives a notification, and the request itself is redirected to Claude Opus 4.8. According to the company, the new classifier detects the targeted method in over 99% of cases.

At the same time, Anthropic acknowledges: "It is likely impossible to make any AI model completely resistant to jailbreaks." The new measures will lead to an increase in false positives during routine programming and debugging tasks. This is a deliberate compromise in exchange for broader access to the model's other capabilities.

A Consensus Framework for Threat Assessment

Anthropic, together with Amazon, Microsoft, Google, and other Project Glasswing partners, has begun developing a common approach to assessing the danger of jailbreaks. The industry still lacks a unified standard for determining the severity of such security bypasses.

The proposed scheme evaluates a jailbreak based on four criteria:

  • how much it expands the attacker's capabilities compared to available tools;
  • how many different malicious tasks the bypass covers;
  • how easily it can be turned into a real attack;
  • how easily other people can find or replicate this method.

For the most dangerous cases, the company promises to immediately launch preliminary protective measures. Anthropic is also forming a round-the-clock team to monitor channels where jailbreak data is published, and is launching a HackerOne program for reporting potential security bypasses of Fable 5.

Access to Mythos 5 has already been restored for some U.S. organizations following government approval on June 26. The company will continue to cooperate with authorities, including pre-release access to protective tools, sharing information about hacks, and dedicated resources for joint research.

My analysis: The situation with Fable 5 and Mythos 5 is a vivid example of how AI regulation in the U.S. is becoming increasingly strict, yet flexible. Anthropic demonstrates a willingness to compromise by implementing complex filtering systems, while acknowledging their imperfections. This sets a precedent for the entire industry: expect other major players to follow suit in the near future, developing their own "consensus frameworks" for threat assessment. Investors should pay attention to companies actively investing in security — this will become a key factor in competitiveness amid tightening regulations.