ARCHIVE
All editions.
Every briefing, fully archived. Five to six new ones every week.
2 briefings. Filtered by ai-policy — clear filter.
- EDITION #310
Two Models From the Same Base Scored 1.5% and 95% on the Same Test | Edition 310
Edition 310 — OpenAI's GPT-5.6-Cyber found two Chrome zero-days. The number that matters measures willingness, not skill.
- EDITION #188
The AI Model Anthropic Refuses to Ship
Mythos is so good at finding software vulnerabilities the White House is drafting FDA-style approval rules. The deregulation era of AI just ended. w