Day August 27, 2026

AI Agents Are Forcing a Surprising US-China Safety Rethink

AI safety concerns are pushing US and Chinese researchers toward limited cooperation as AI agents become more capable and harder to control.

Gemini Notebook now lets users query purchased books from Google Play Books, turning them into AI sources for summaries, plans and more.

Barret Zoph joins Google as vice president of research after brief OpenAI return, underscoring the AI talent shuffle around Gemini.

Anthropic unveiled an AI agents hardware standard to help labs and factories use robots and instruments more safely.

A Georgia officer resigned after alleged Flock misuse to track an ex and another cop, exposing broader risks in police surveillance.

OpenAI, Google and 100+ companies warn on AI cyber threats, urging governments and industry to build stronger defenses fast.

OpenAI is testing a persistent AI agent in Codex, aiming for proactive, cross-session help while navigating major safety risks.

Jensen Huang said Nvidia has reached AGI for some tasks, then called the AGI debate senseless as he shifted focus to useful AI and tokens.

Google’s AI Mode travel update adds flight price tracking, hotel booking help, and points-or-miles pricing in a big search upgrade.

Google is tightening Android app rules as the Android memory crunch from AI chip demand threatens app performance and device reliability.