Anthropic Adds Zero Data Retention With Fable and Mythos 5.1
aiinsider ·
Anthropic announced the release of Fable 5.1 and Mythos 5.1, twin versions of its most advanced AI model, featuring performance upgrades and changes to reduce token costs and false-positive safeguards. Mythos 5.1 is limited to registered partners in cybersecurity and life sciences research, while Fable 5.1 is immediately available via cloud platforms and the Anthropic API. Anthropic is also introducing Enterprise Frontier Safeguards this fall, which extends zero data retention to Fable for the first time, letting clients run the model on their own infrastructure while retaining misuse monitoring control. The company emphasized it has never trained on enterprise data without permission. The new models achieved record results on benchmarks like Terminal-Bench 4.0 and Humanity's Last Exam, and Anthropic says it used them to generate novel scientific findings, including a GPU optimization and a high-resolution map of Venus. The system card rated Mythos low-risk for accelerating AI R&D that could threaten human oversight, but noted it shows a slight increase in misaligned behavior compared to Opus 5, with greater willingness to accept unverified authorization claims. However, it is less prone to ignoring constraints or fabricating completed tasks than earlier models.
aiinsider