#Anthropic

Anthropic has announced Claude Fable 5, its first publicly available Mythos-class AI model, offering improved performance across software engineering, research, vision, and complex knowledge tasks. To reduce potential misuse, Anthropic has implemented safeguards that redirect certain cybersecurity, chemistry, and biology-related queries to the older Claude Opus 4.8 model. Alongside Fable 5, Anthropic introduced Claude Mythos 5 for select cyberdefenders and infrastructure providers through its Project Glasswing program. Anthropic has called for a coordinated global pause on the development of the world's most advanced AI systems, warning that future models could become difficult for humans to control. Anthropic said such a pause would only be effective if major AI developers and governments, particularly the United States and China, agreed to participate under a verifiable international framework. Anthropic warned this could eventually lead to "recursive self-improvement," where AI systems become capable of improving themselves with limited human involvement. Anthropic acknowledged that achieving a global agreement would be difficult due to competitive and geopolitical pressures, as companies and governments continue to invest heavily in AI development. Key points Anthropic proposes a temporary pause on frontier AI development Calls for international coordination involving the US and China Warns about potential loss of human oversight in future AI systems Highlights risks of recursive self-improvement Says AI is increasingly helping develop newer AI models Read more in our articles including "Anthropic launches Claude Fable 5" and "Anthropic proposes global pause on advanced AI development".