Stella Biderman (@BlancheMinerva) / X

Stella Biderman

14.3K posts

Stella Biderman

@BlancheMinerva

Ensuring that tech companies don't have a monopoly on being able to do research on cutting edge AI @AiEleuther. She/her

Joined May 2019

Pinned
Stella Biderman
@BlancheMinerva
Aug 7
If you follow me and haven’t listened to this talk in its entirety, doing so is more important than whatever it is you are currently doing. There’s a lot of cybersec jargon at times but I don’t think you need to know anything about security to understand what’s important.
Eric Wallace
@Eric_Wallace_
Aug 6
Yesterday, my OpenAI collaborator and I gave a detailed talk on the Huggingface incident, our models creating "the message board", model misalignment, and more. youtube.com/watch?v=87DyyM… I hope it can answer a lot of the questions folks have, and we will release a full detailed
Stella Biderman
@BlancheMinerva
2h
Saving this for the next time a frontier lab starts talking about "stealing chain of thoughts" as a illegal offensive cyber operation carried out against them.
Alexander Panfilov
@kotekjedi_ml
9h
Replying to @kotekjedi_ml
Some background: In May, @matthew_d_green found that encrypted reasoning could be replayed outside its original context, and reported it to the labs (blog.cryptographyengineering.com/2026/05/29/foo…). The labs said that "they don’t see any security implications in side channels or replays". In our
Stella Biderman
@BlancheMinerva
2h
And here I was thinking that Claude was starting to get me, lol
Stella Biderman
@BlancheMinerva
Aug 10
Hmmm so it looks like Claude was asked to put someone on a waitlist and decided to hack their system instead of saying “sorry, the waitlist isn’t open yet”? I’d love to know more about how much pushing this required, what the prompts were, etc.
How a simple request for AI to book a gym class exposed a major threat
From abc.net.au
Stella Biderman
@BlancheMinerva
Aug 9
Weird, I thought the more important point is that OpenAI has been ignoring computer security best practices and their own published work on monitoring AI systems?
Dean W. Ball
@deanwball
Aug 8
It is true that the hugging face incident is an example of a malicious, emergent digital ecology of machine intelligence. But the more important point is that digital ecologies of machine intelligence can be grown! Yes, we accidentally made a weed. And yes, nasty actors will make