Agent Evaluation Gaming: METR and Redwood Say the Agents Went After the Scorer, Not the Benchmark
We have written about the OpenAI and Hugging Face incident twice, and both times on the framing that a set of evaluation…
Cyber security is the practice of protecting computer systems, networks, and data from digital attack.
We have written about the OpenAI and Hugging Face incident twice, and both times on the framing that a set of evaluation…
Package registry security is not a thing most teams have a plan for. The registry is plumbing. It caches dependencies, it sits…
Local ai agent security has a founding assumption that almost nobody states out loud, and CVE-2026-65105 is what happens when it fails.…
The cra reporting obligations under the EU Cyber Resilience Act take effect on September 11, 2026. From that date, manufacturers of products…
Two things happened on August 22, 2026, and the pairing is the story. OpenAI asked California to make a law stricter, and…
Microsoft is ending phone-based multi-factor authentication in Entra ID, and the first of two deadlines arrives at the start of September. If…
Agentic browser security comes down to one substitution, and most coverage misses it. The web has a single boundary everything else is…
GPT-5.6-Cyber is a frontier model with its safety refusals deliberately turned down, released by OpenAI on August 10, 2026 to an approved…
WordPress 7.0.3 landed on August 6, 2026 carrying a fix for something that sounds worse than almost anything WordPress has shipped a…
AI evaluation containment has now failed publicly at three frontier labs, and the same testing vendor is named in three of the…