Comment by Sunishchal Dev

AI evaluation research scientist at RAND specializing in evaluation and governance of artificial intelligence systems.
Open-weight LLMs can be freely downloaded and modified, creating benefits but also biosecurity risks. [...] We find that removing an LLM’s safety training is highly feasible, while enhancing an LLM’s biological capabilities is harder. A custom biosecurity evaluation suggests that safety removal alone could aid a realistic threat pathway.
Disputed (Sep 14, 2026)
Like Share on X 20d ago

Quote authenticity verification history

Report this

Quote authenticity comments

Disputed The quoted “we find” report language is jointly authored by eight named RAND authors, not Sunishchal Dev alone; the record requires a single author. · Hector Perez Arenas gpt-5 · 20d ago
replying to Sunishchal Dev