Claude output watermarks, a Riemann bound, Meta's open weights - August 11, 2026

Chapters

1. 0:00

   Intro

2. 0:20

   Anthropic starts watermarking every Claude output

   https://support.claude.com/en/articles/16266773-how-claude-marks-ai-generated-content
3. 1:14

   Claude pushes a Riemann hypothesis bound from 41 to 67 percent

   https://www.anthropic.com/research/riemann-zeta
4. 2:11

   Meta returns to open weights with Muse Glimmer

   https://simonwillison.net/2026/Aug/10/introducing-muse-glimmer/
5. 2:58

   Dan Luu tests whether dynamic languages really cost fewer tokens

   https://danluu.com/pl-tokens/
6. 3:46

   OpenAI ships a cyber-trained model and splits Daybreak in two

   https://techcrunch.com/2026/08/10/as-ai-led-attacks-multiply-openai-launches-a-new-cyber-model/
7. 4:35

   An agent broke into a gym booking system nobody asked it to test

   https://techcrunch.com/2026/08/10/tech-industry-is-buzzing-after-a-claude-agent-hacked-into-a-gym/
8. 5:19

   A Metabase zero-day turns into a three-company breach

   https://www.helpnetsecurity.com/2026/08/10/metabase-zero-day-framework-tally-kilo-code/
9. 6:10

   The Ceva Logistics breach ripples into banks, retailers and Steam

   https://techcrunch.com/2026/08/10/a-data-breach-at-shipping-giant-ceva-logistics-is-rippling-across-banks-retailers-steam-gamers-and-beyond/
10. 7:02

    AI agents found 84 flaws in mobile core network software

    https://www.helpnetsecurity.com/2026/08/11/5g-core-network-vulnerabilities-research/
11. 7:51

    Live facial recognition reaches the London Underground

    https://idtechwire.com/british-transport-police-expands-facial-recognition-trial-to-london-underground/
12. 8:36

    Sign-off