Edition Summary (20 Pages)

AI safety and control dominated the edition as OpenAI and Anthropic leaders urged global cooperation and stronger safeguards at the UN Security Council, while separate reports highlighted autonomous agent security incidents involving OpenAI and Google Gemini. Competition among major model builders also intensified, with new OpenAI, Anthropic and Google releases emphasising lower costs, stronger performance and expanded voice capabilities. Agentic AI moved deeper into commerce and enterprise workflows through Meta’s Muse, Amazon seller automation, Salesforce, ServiceNow, Workday and Grok integrations. Investment remained substantial, including Microsoft’s planned Gulf cloud and AI expansion and wider sovereign AI infrastructure initiatives, while research coverage showed AI agents increasingly being used for scientific discovery, biotechnology and automated research.