Legal Advocates for Safe Science and Technology is dedicated to using legal advocacy to make advances in science and technology safer for people and the planet.
New York, NY
Joined December 2025
- Tweets93
- Following149
- Followers407
- Likes53
Pinned Tweet
We're suing OpenAI over its hack of Hugging Face.
AI companies are building agents that are autonomously making decisions, taking actions, and accessing systems without human direction at every step.
The risks highlighted by OpenAI’s hack of Hugging Face are only going to multiply as these agents get even more powerful. 🧵
A company whose technology threatens catastrophic harm should not be less accountable for S-1 misstatements than almost every other company that IPOs.
1
8
386
Earlier this year the White House announced its voluntary AI testing framework through which it would receive early access to frontier AI companies' most advanced models before their public release.
On Friday, in response to our FOIA lawsuit, we were told by the government that no part of the 7-page framework will be disclosed to the public.
LASST will continue fighting for the framework’s public release.
Read more on our Substack and in @theinformation by @leomschwartz:
Read more here:
lasstorg.substack.com/p/the-…
And here: theinformation.com/briefings…
And from @protctdemocracy, who has filed a similar lawsuit here:
protectdemocracy.org/work/un…
5
219
An update on our lawsuit seeking the Pentagon’s AI contracts:
We have received the most recent version of its OpenAI contract, from April 2026. The 16-page contract includes the limits on domestic surveillance and independent autonomous killing that OpenAI announced in March, though we’ve questioned whether they provide the protections OpenAI claims.
The Pentagon also produced a 1.5-page memorandum of understanding with Reflection AI, providing that Reflection AI’s technology will be deployed on its classified networks.
The agreement promises that Reflection will not attempt to restrict DOD’s use of its technology beyond what the law requires.
1
6
139
We asked for these contracts on behalf of @theintercept after the Pentagon and Anthropic clashed over limits on military AI. When the Pentagon missed its deadline, we sued.
All documents: lasst.org/transparency/dod-a…
2
5
112
As AI companies deploy systems that act without human direction at every step, the question of who is responsible when they cause harm is getting more urgent. California law is clear on one point: companies can't avoid liability simply by arguing that “an AI did it.”
LASST Programs Director Vivian Dong spoke to @BostonGlobe's @HiawathaBray about our lawsuit against OpenAI over its agents' unauthorized access to Hugging Face.
(Link in comments)
Arkansas has enacted new laws protecting survivors of AI-generated sexual deepfakes. What do these protections mean in practice?
Tomorrow, LASST is partnering with @heatinitiative on Justice in the Age of AI, a survivor open mic and community event in Fayetteville, Arkansas, exploring legal pathways to accountability.
Free to attend, all welcome - RSVP below ⤵️
1
6
161
