The Trump administration has partially lifted its export ban on Anthropic's most advanced AI models, Mythos 5 and Fable...

The Trump administration has partially lifted its export ban on Anthropic's most advanced AI models, Mythos 5 and Fable 5, allowing limited release to trusted US organizations (Politico, NPR, WSJ). This follows months of restrictions that had kept these models from key customers and fueled the rise of Chinese and Japanese rivals claiming parity (TechCrunch). The partial carve-out signals a thaw in US-Anthropic relations and positions the company to reclaim enterprise momentum. Concurrently, Anthropic is expanding internationally with a hiring spree in Europe (Munich, Paris, London, Zurich) (Greenhouse job postings), and Austria is lobbying the EU to host Anthropic operations (Bloomberg). On the product front, Anthropic's Claude Fable 5 leads the LMSYS Chatbot Arena with an ELO of 1508 (LMSYS), and the new Claude Tag Slack integration has drawn both excitement and concern from Salesforce employees over potential cannibalization (The Information). However, security researchers at Mozilla's 0din demonstrated a multi-step exploit for Claude Code (Tom's Hardware), and China's underground Claude access economy persists (Wired). Overall, the regulatory easing is the dominant signal, enabling monetization of Anthropic's frontier models and reinforcing its competitive position against OpenAI and Chinese LLMs.