Anthropic releases Claude Fable 5, a 'Mythos-class' AI model with safeguards (www.businessinsider.com)

Anthropic has announced the release of Claude Fable 5, a 'Mythos-class' AI model, emphasizing its robust safeguards designed to prevent potential misuse in sensitive areas like cybersecurity. While Fable 5's capabilities are groundbreaking, the model will redirect certain queries to the less powerful Claude Opus 4.8 to mitigate risks of damaging misuse. This release aligns with Anthropic's ongoing Project Glasswing, granting access to a select group of cybersecurity professionals through the Claude Mythos 5 model, which offers advanced features without some of the restrictions present in Fable 5. This development is particularly significant as it comes amidst heightened scrutiny of AI technologies and their implications for national security and data protection. The advance of Mythos-class models has amplified concerns about the potential for AI to exploit vulnerabilities in software systems, prompting governmental responses such as President Trump’s recent executive order allowing AI companies to voluntarily provide access to new AI models before their public release. The competitive landscape is also shifting, with Anthropic and OpenAI both moving towards IPOs, indicating a crucial moment for investment and regulatory discussions surrounding AI advancements.