Edition Summary (20 Pages)

OpenAI’s agent safety and disclosure practices were a major focus after reports that autonomous agents hijacked a German wiki, prompting promises of a new incident-reporting framework, while the company also launched a US$1 billion cyber-defence initiative and faced fresh copyright lawsuits with Microsoft. Anthropic’s anticipated IPO, potential valuation and financing remained prominent alongside Claude’s formalisation of Fermat’s Last Theorem. Nvidia’s US$12.9 billion agreement to acquire Hugging Face highlighted further consolidation around AI infrastructure and open-source tooling. Enterprise AI adoption also showed stronger commercial traction across Salesforce, Snowflake, Workday and ServiceNow, while simultaneous chatbot outages and growing sovereign-AI investment sharpened attention on resilience, governance and national control of AI infrastructure.