Insights

BNY Mellon cut client research from two days to ten minutes. Siemens saved energy in Swiss buildings with closed-loop AI. OpenAI started testing ads in ChatGPT. Microsoft shipped planning agents. New benchmarks show AI still fails most real work. This is where business stands now.

Shopify ships agentic shopping that sells inside chats. Apple picks Google Gemini to upgrade Siri. Anthropic lets Claude “Cowork” on your files. Microsoft responds to data-center power and water pressure. Regulators tighten rules on AI misuse and copyright.

Retailers rolled out AI agents that complete purchases. Nvidia unveiled chips built for multi-agent systems. OpenAI launched a health product tied to medical records. Swiss hospitals began testing local medical AI. At the same time, regulators moved to slow things down.

The U.S. military rolled out AI agents at scale, Tesco signed a multi-year deal with Mistral AI, Disney embedded generative AI across its operations, regulators delayed key AI rules, and researchers used AI to discover hundreds of thousands of new materials.

Zoom turns its AI into an active workplace agent. OpenAI ships a new coding model. Google pushes Gemini deeper into products. Banks, engineers, and researchers deploy AI in live systems, while regulators and Swiss startups respond in real time.

OpenAI shipped GPT-5.2 as Google rolled out agent connectors and a Gemini research agent. Microsoft pledged $23B for AI infrastructure. Virgin Atlantic and Swiss Re showed real deployments, while EU and US policy shifted and new chips and super-networks pushed compute forward.