News

What changed, what it means, and whether it matters.

AI Agents Can Breach Corporate Test Networks Autonomously Vals AI found that autonomous agents can navigate simulated corporate networks and reach admin-level access without human help.
Why AI Coding Benchmarks Fail in Real-World Tasks Can Smartwatches Flag High Blood Pressure Without a Cuff? Nvidia’s Flora Cuts Robot Labeling Costs Claude Duet: Your AI Assistant That Works in the Background Google Cloud Adds OpenAI, Joining Gemini and Claude OpenAI Says Frontier Training Now Runs Without Internet Access Taste-Bench shows AI agents still struggle with long-term decisions. Amp expands its AI agent beyond coding with OpenAI models Gemini 4 Argon builds robot training worlds from any image or video. Third Moment says OpenAI’s stack runs 10,000 agents for 48 hours. Odyssey-3 launches free public demo for interactive AI worlds Do AI Agent Teams Really Work Better? Benchmarks Say Usually Not Nano Banana 2.1 cuts Google's image generation prices nearly in half Gemini 4 Argon: What Google's New Model Means for Creators Gemini Accidentally Hacked Real Companies in a Security Test: What We Know