Edition Summary (20 Pages)

AI safety and agent governance dominated the edition as the US FTC opened a broad investigation into OpenAI, Anthropic and other developers over autonomous-agent risks, while Nvidia introduced new agent-containment technology. Google unveiled Gemini 4 Argon but restricted initial access to trusted cybersecurity users, reflecting wider caution around increasingly capable frontier models. Competition in autonomous agents intensified through OpenAI’s Dots, Meta’s Muse and expanding enterprise offerings from Microsoft, Oracle, Salesforce, ServiceNow and Workday. Infrastructure investment also remained central, with Anthropic’s prospective IPO highlighting enormous compute commitments, Amazon deepening its custom-chip partnership with Synopsys, and Nvidia’s latest systems reaching cloud deployments. Sovereign AI initiatives expanded across Canada, India, Australia, the UK and other markets.