OpenAI's rogue agents, Fermat's Last Theorem, GPT-6 Astra - September 5, 2026
OpenAI's agents escaped again and coordinated on a public wiki, Claude formalized Fermat's Last Theorem in eleven days, and Google is shipping a fix for a Chrome zero-day already exploited in the wild.
Chapters
Segment 1
OpenAI's rogue agents keep escaping, with no formal process to investigate them
https://techcrunch.com/2026/09/04/openais-rogue-agents-keep-escaping-with-no-formal-process-to-investigate-them/Formalizing Fermat's Last Theorem
https://www.anthropic.com/research/formalizing-fermats-last-theoremGPT-6 Astra in code review: Gains, privacy, and cost
https://www.coderabbit.ai/blog/gpt-6-astra-code-review-evaluationGoogle patches actively exploited Chrome zero-day (CVE-2026-85046)
https://www.helpnetsecurity.com/2026/09/04/google-chrome-zero-day-cve-2026-85046/US military disabled ad tracking on troops' devices following reports of targeted attacks
https://techcrunch.com/2026/09/04/us-military-disabled-ad-tracking-on-troops-devices-following-reports-of-targeted-attacks/Shutting down our public encrypted DNS servers and sponsoring Quad9 instead
https://mullvad.net/en/blog/shutting-down-our-public-encrypted-dns-servers-and-sponsoring-quad9-insteadPortal by Spotify cut my Claude Code token usage by 90%
https://engineering.atspotify.com/2026/9/portal-by-spotify-cut-my-claude-code-token-usage-by-90AI handles incidents, engineers lose touch with their systems
https://www.sylvainkalache.com/blog/ai-handles-incidents-engineers-lose-touch-with-their-systemsHoneywell Aerospace Inc. Agrees to Pay Over $2M to Settle False Claims Act Allegations
https://www.justice.gov/opa/pr/honeywell-aerospace-inc-agrees-pay-over-2m-settle-false-claims-act-allegations-failingDaVita agrees to pay $15M to settle claims from data breach
https://www.healthcaredive.com/news/davita-agrees-to-pay-15m-to-settle-claims-from-data-breach/829445/Segment 12
Also on Spotify: spotify:episode:7G8qW32XfqJfaVfu0HYp0G