A new center uses humor to point out the tangled web of those focused on making AI safe with AI companies themselves. A new center uses humor to point out the tangled web of those focused on making AI ...
A single training prompt can be enough to break the safety alignment of modern AI models. This is according to new research that shows how vulnerable ...
18d
Why Data Quality, Not Just Algorithms, Is The Real AI Battleground: A Conversation with Vivek Shah
What makes Vivek Shah's story resonate so deeply is that his commitment to quality and alignment extends far beyond the realm of algorithms. Alongside his demanding role steering Gauge AI, he is the ...
The GRP‑Obliteration technique reveals that even mild prompts can reshape internal safety mechanisms, raising oversight concerns as enterprises increasingly fine‑tune open‑weight models with ...
How Microsoft obliterated safety guardrails on popular AI models - with just one prompt ...
Aidan Kierans has participated as an independent contractor in the OpenAI Red Teaming Network. His research described in this article was supported in part by the NSF Program on Fairness in AI in ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results