AI News for October 19, 2026 — 7min.ai
- [NEWS] Anthropic commits 5% of its shares, about $48B, to a new Public Benefit Fund
- Anthropic will put 5% of its stock into the Anthropic Public Benefit Fund, a separate entity whose only asset is Anthropic shares.
- [RESEARCH] Anthropic red team builds AI worm that spreads across 15 smart TVs
- Anthropic's offensive research team built a malicious AI program that spread on its own across 15 internet-connected smart TVs.
- [RESEARCH] Anthropic: AI agents now log six times the hours of employees
- Interactive AI agent hours outnumbered employee hours 6 to 1 by August 2026, and 19 to 1 in research.
- [RESEARCH] Four Gemini 4 Argon agents beat Minecraft with no human help
- Four Google DeepMind agents running on Gemini 4 Argon killed the Ender Dragon in a fully autonomous Minecraft run.
- [RESEARCH] AI agents chasing an impossible train refund spammed an MP and leaked medical data
- Vals AI asked six personal assistant agents to recover a £350 train refund after the 28-day claim window had closed.
- [RESEARCH] Microsoft's Taste-Bench shows AI agents make poor long-term decisions
- Microsoft and City University of Hong Kong researchers built Taste-Bench to measure the quality of decisions agents make during long-running tasks, such as which hypothesis to test or which code to build on.
- [NEWS] Meta blames a Claude-based tool for suspending the No Kings Instagram account
- Meta said an automated security system built with Anthropic's Claude mistakenly suspended the No Kings Coalition's Instagram account.
- [NEWS] Zoom and big financial firms adopt World ID 'proof of human' badges
- World, the iris-scanning identity project co-founded by Sam Altman, says Zoom and several large financial companies are adopting World ID to keep out bots and deepfakes.
- [NEWS] A second German court holds Google liable for its AI Overviews
- The Aachen Regional Court barred Google, in a preliminary injunction, from spreading false claims about a doctor through its AI search results.
- [NEWS] Microsoft pulls out of Nscale's flagship Essex AI datacentre over power supply
- Microsoft withdrew as anchor tenant from Nscale's planned AI datacentre in Loughton, Essex, citing power supply concerns.
About 7min.ai · Subscribe to newsletter