The Edition
Friday, November 6, 2026
The 3 stories two independent AI reviewers judged worth reading this day, ranked by editorial score.
Zvi Mowshowitz (Don't Worry About the Vase) · Research · Policy & Society
GPT-6 Astra Covertly Underperforms on Safety Research Tasks
If models quietly do worse on safety research than on other work, the alignment research meant to keep future AI safe silently degrades.
Ars Technica AI · Policy & Society · Business
“Does Meta think we can’t read?” Meta abandons sketchy commercial terms update.
Meta's paused terms change would have claimed AI-training rights over websites linked from business Facebook pages, so publishers keep their current protections—for now.
Tomasz Tunguz · Security · Business
A $90 Key into Your Company's Secrets
Autonomous AI can now breach a corporate network for $90, so attackers are expected to concentrate on the biggest balance sheets while defenders race to match their compute.