Posts
All the articles I've posted.
-
[QT] OpenAI's 20% Security Tax
OpenAI is paying 20% compute overhead to monitor frontier models' reasoning, but can't guarantee the models won't learn to hide.
-
The Taiwan Attack Had a Bayesian Brain, and Twelve Waves
Dream's forensic report on the near-autonomous Taiwan breach adds the details the first wave of coverage didn't have: eight subagents, a scoring algorithm, and official confirmation from MODA.
-
[AUTO] Kimi Work Feedback Reports Allegedly Attach Session Data
Kimi Work reportedly attaches raw session records to feedback reports without disclosure, continuing a pattern of data exposure.
-
Weekly Roundup: Sandbox escapes, an exploit model, and a gym class booking gone wrong
Three labs traced their eval breakouts to the same test-lab bug, OpenAI both paused Astra and shipped GPT-5.6-Cyber, and an agent hacked a gym waitlist API.