Anthropic has resumed global availability of its Claude Fable 5 artificial intelligence model after U.S. export controls were lifted. The relaunch restores access to what the company calls its first Mythos-class AI, designed to perform complex tasks in software engineering, vision, scientific research, and extended autonomous operations.
Initially launched in early June 2026, Fable 5 demonstrated superior capabilities compared to previous Anthropic models, excelling in multi-day, intricate projects. However, concerns over misuse—primarily in cybersecurity—prompted U.S. authorities to impose export restrictions shortly after the debut. This forced Anthropic to temporarily halt international access while improving security safeguards.
The company responded by deploying a new classifier designed to block high-risk techniques in over 99% of cases, addressing the government's concerns. The updated system now enables safe, widespread usage of the model. Alongside Fable 5, Mythos 5, a variant with fewer restrictions but the same core functions, is again available to certain U.S. agencies focused on critical infrastructure defense through Project Glasswing.
Anthropic allows eligible subscribers—including Pro, Max, Team, and selected Enterprise plans—to use Fable 5 for up to half of their weekly AI usage through early July 2026, after which usage will convert to credit-based access. The model is available across several platforms, including Claude.ai, the Claude Platform, Claude Code, and Claude Cowork, and is rolling out on leading cloud providers such as AWS, Google Cloud, and Microsoft Foundry.
Apple users benefit from native compatibility, as Fable 5 operates smoothly on Mac computers via the official Claude Desktop app for macOS, alongside iOS and iPadOS versions. This integration allows developers and professionals using Apple Silicon hardware to seamlessly incorporate advanced AI into their workflows.
Anthropic has highlighted its strengthened collaboration with the U.S. government, involving pre-release testing and research partnerships. This relationship illustrates an emerging framework that balances innovation with regulatory responsibility for future AI deployments.
Early adopters praise Fable 5's autonomy and precision in handling long-form and technically complex assignments, marking it as a significant productivity tool for software development, research, and creative endeavors. The model’s reinstatement signals notable progress in expanding sophisticated AI access while managing associated risks.