Anthropic is securing six years of AI compute from Volta, a new Nvidia-backed cloud startup, deepening infrastructure ties and concentration risk for builders and buyers.

NVIDIA and partners launched the Open Secure AI Alliance to build open tools for securing AI agents across multi-vendor clouds and enterprise stacks.

A German consortium released Soofi S, an open 30B model trained on Deutsche Telekom cloud, claiming top fully open results in German and English.

Vercel CEO Guillermo Rauch says enterprises are moving from single-lab AI bets to modular agent stacks, reshaping infrastructure choices.

Mistral AI is gaining new attention as Europe pushes AI sovereignty, highlighting its enterprise-heavy strategy, open-weight models, and cloud ambitions.

Google has capped Meta's use of its Gemini AI models after the social media giant demanded more computing capacity than Google could supply.

The EU unveiled a technology sovereignty push focused on cloud, AI capacity, chips, open source, and reduced reliance on US tech.

Reports about OpenAI’s compute strategy and Amazon access sharpened attention on AI cloud dependence and costs.

Amazon began offering OpenAI products on AWS after Microsoft and OpenAI loosened their exclusive cloud relationship.

Microsoft's new MAI-Image-2-Efficient model cuts image generation costs by 41%, targeting high-volume commercial use cases like product shots and marketing assets.

Following the Pentagon's supply-chain risk designation of Anthropic, cloud giants Microsoft, Google, and Amazon have confirmed that Claude remains available to all their customers except those using it directly under Department of Defense contracts, limiting the practical impact of the designation.