AI agents are getting better, Microsoft is putting enterprise integration at the center, and Mecklenburg-Vorpommern is betting on open source. Plus: new research on interpretability and inference.
Anthropic, Google, and new research shape the AI day: from science workflows and faster image models to safety issues and benchmarks.
Anthropic, Google, and the chip industry are sending strong signals today: new models, more infrastructure, and new risks for AI workflows.
Samsung and SK Hynix are expanding massively, Claude Code stumbles over manipulated repos, Apple patches AI risks, and Princeton tests startup AIs.
Today in AI Radar: new research on reward models, agent benchmarks, time-series foundation models, and why companies are suddenly rethinking AI costs.
GPT-5.6 launches despite a U.S. regulatory dispute, METR reports test cheating, plus new research on Terraform fixes, multi-agent systems, and AI bubbles.
OpenAI launches GPT-5.6, chatbots continue to show political bias, and a startup saves millions with Deepseek. Plus new benchmarks and research.
New research on transfer and forecasting, bias in chatbots, Figma under AI pressure, and China’s low-cost models putting the West under pressure.
Affordable Chinese AI, new inference chips, Cursor plans, and Figma updates show: the AI market is getting tougher, faster, and cheaper.
AI infrastructure is booming, Adobe is bringing agents into creative apps, DeepMind is sharpening security rules, and new research is adding fresh context.