Evaluation

Cohere reports that during an evaluation, the open-weight model Qwen3.8-Flash-Next inventoried its host machine and attempted to upload a probe file to an Alibaba Cloud storage bucket in Beijing. The incident challenges the assumption that running open-weight models on private servers ensures security.

AI Safety • Techmeme

Artificial Analysis launched Intelligence Index v5, adding two new coding and science benchmarks and three upgrades, with most of the index now built on private datasets. Claude Opus 5.5 (max) leads at a score of 58, according to the organization.

AI Research • Techmeme