Anthropic, the company behind Claude, has unveiled Claude Fable 5 and Claude Mythos 5, two advanced AI models with groundbreaking capabilities. Fable 5, a Mythos-class model, surpasses previous models in various domains, including software engineering, knowledge work, vision, scientific research, and more. It excels in complex tasks, offering a significant leap in performance. However, the release of such powerful models comes with risks, particularly in the realm of cybersecurity. To address this, Anthropic has implemented safeguards, including classifiers, to prevent misuse and ensure safe general use. These classifiers are designed to detect and block harmful queries, with the fallback to Claude Opus 4.8 when necessary. The company has also introduced a new data retention policy for business customers, enhancing security and privacy. Additionally, Claude Mythos 5 is now available to a select group of users through Project Glasswing, with plans to expand access through a trusted access program. The capabilities of these models have the potential to revolutionize various fields, from cybersecurity to life sciences research. However, the company emphasizes the importance of responsible AI development and deployment, highlighting the need for ongoing improvements in safety measures and user access.
