News archive · 2026-09-18

Daily AI News and Tools

4 local AI news items from regional sources.

Anthropic relaunches Claude Code Projects for multi-agent cloud work

Anthropic has relaunched the Projects feature in Claude Code, letting users run multiple agents under one project with shared memory and goals. The shift moves AI coding from isolated chats toward coordinated multi-agent workflows, changing how developers split tasks and manage context.

开发者生态Source: The Verge AI

Podcast examines China's open-weight models, Nvidia and the AI security dilemma

Part two of AI: Reset to Zero episode 1212 discusses the so-called AI security dilemma, touching on China's open-weight models and Nvidia. The episode frames open-weight competition and chip supply within the same security debate, a framing that increasingly shapes model release strategies and cross...

模型开源Source: AI: Reset to Zero

The AI superintelligence slowdown: safety shifts the pace at OpenAI and Anthropic

The Verge reports signs of a slowdown in the push toward AI superintelligence, touching on the safety posture of OpenAI and Anthropic, after years of 'move fast and break things' in tech. The shift from pure speed to safety and pacing affects model release timelines and product iteration, changing h...

政策安全Source: The Verge AI

UN turns to Google to make global data ready for AI agents

TechCrunch reports the United Nations is working with Google to make its global data ready for AI agents. The move marks an international body restructuring public data for agent-readable use rather than human search alone, which could affect data access, permission design and compliance boundaries ...

开发者生态Source: TechCrunch