[Thread] A roundup of recent announcements from researchers, businesses, and academic institutions opposing California's AI safety bill
the real experts — agree the risks are real. We all urge you to pass it into law. Titus Wu / @tituswu100 : Pressure is intensifying on #SB1047, California's controversial #AI safety bill, as lawmakers near the end of #CALeg session. It's also getting some new scrutiny from the congressional level. Story w/ @omaseddiq https://news.bloomberglaw.com/ ... Prof. Anima Anandkumar / @animaanandkumar : In my @TEDTalks I talk about the immense impact AI is having in scientific domains: faster and better modeling of extreme weather events, getting closer to realizing the dream of nuclear fusion, and design of medical devices that cuts down infections by 100x...@Scott_Wiener Please do not just focus on worst-case doomsay outcomes that can kill innovation in academia. AI has always been an open-source revolution. Let us preserve it. Anjney Midha / @anjneymidha : We've submitted a 14 page response to @Scott_Wiener to set the record straight once and for all on the dangerous impacts of SB 1047 on AI safety, open source and Little Tech. Pls help amplify. [image] Dan Hendrycks / @danhendrycks : Notion co-founder @simonlast writes “SB 1047 strikes a balance between protecting public safety from such harms and supporting innovation, focusing on common sense safety requirements for the few companies developing the most powerful AI systems.” https://www.latimes.com/... @martin_casado : .@drfeifei one of California's true AI pioneers lays out exactly how SB 1047 will harm startups, open source, and academic research. https://fortune.com/... https://x.com/... @martin_casado : .@russellwald and @AndrewYNg from Stanford, @vishalmisra from Columbia, and Ion Stoica from Berkeley have also articulated the damage the bill would do to their research, the AI industry and to the state of California. [image] Rob Saker / @robsaker : California SB 1047 is such a horribly written, misguided bill. It will do nothing improve the safety of AI. It will move AI development out of California and enrich a few very large tech companies. It is an egregious example of regulatory capture. https://reason.org/... @aftfuture : An important summary of the emerging, future-focused coalition against a bill that would harm AI startups across the country @abcampbell : @NathanpmYoung ... Guess it depends on what your definition of “overwhelming” is. Seen a lot of polls and as many accusations of push polling by special interest groups. Not sure I know what I believe anymore, which is why I wouldn't call it “overwhelmingly popular.” Dean W. Ball / @deanwball : I felt rather lonely for several months after I first wrote about SB 1047 in February. I don't feel lonely anymore! The AI field around the world is watching now, and they believe, correctly, that it is a bellwether for America overall. Will the USA succumb to fear? Brian Chau / @psychosort : @Scott_Wiener isn't backing down from his attempt to ban open source AI in exchange for money from the EA doomsday cult. @martin_casado has an excellent thread tracking the unified opposition from every part of California Andrew Ng / @andrewyng : @martin_casado has a nice thread summarizing some of these...If anything, by hurting open source and thus hampering researchers' ability to study cutting-edge models to identify problems, I believe SB1047 is more likely to make AI less safe. Senator Scott Wiener / @scott_wiener : Contrary to what the loudest voices say, SB 1047 began as a grassroots effort with help from folks in SF's hacker houses. It's overwhelmingly popular among Californians, including most tech workers. Today, @NotionHQ's SF-based cofounder @simonlast is out with an op-ed in support [image] Russell Wald / @russellwald : #SB1047 does not offer the regulation California needs and instead attempts combat unsubstantiated X risk while not focusing on the immediate harms of AI. CA needs thoughtful AI regulation that doesn't harm academia's narrowing window in AI. Michael Saxon / @m2saxon : Please read and consider signing our open letter AGAINST the California AI safety bill (SB 1047). It will negatively impact open source model releases, harm reproducibility, and entrench big APIs. We plan to release it before the cmte hearing in 2 days. https://forms.gle/... [image] Noah Smith / @noahpinion : Yes, @Scott_Wiener's AI regulation bill is a misfire. It isn't going to increase AI safety, and it could hurt California tax revenues quite a bit. It should be quietly withdrawn.
Context & Ripple Effects
The opposition is coalescing around a common implementation argument: broad obligations could constrain open-source releases, reproducible research, and smaller developers more than the largest model providers. That closely follows warnings that the measure could harm developers, public-sector users, and academia in an earlier critique of the bill's impact.
The debate is not a simple safety-versus-no-safety split: Dan Hendrycks had described both the bill's rationale and the backlash around it in a prior discussion of SB 1047's supporters and critics. The new round of academic and business objections makes the practical design of state-level AI rules the central point of contention.
First-order effects
- SB 1047's backers and California lawmakers face a more visible, multi-constituency opposition case focused on research access, open-source distribution, and startup compliance.
- Researchers and smaller AI developers gain public arguments to use against the bill, while large API providers are cast by opponents as comparatively better positioned to absorb its requirements.
Second-order effects
- The dispute raises pressure on proponents to show that safeguards can be applied without making open models and academic work less accessible or concentrating advantage among established providers.
- It strengthens the case for locating AI safety rules at the federal level, an argument OpenAI later made in its letter urging a federal approach, rather than letting California set a de facto template alone.
Third-order effects
- If state-level AI rules repeatedly become proxy fights over open-source access and innovation, AI governance may shift toward narrower, capability- or deployment-specific obligations rather than broad developer duties.
- The durable fault line is likely to be who can credibly claim both safety legitimacy and support for an open innovation ecosystem; coalition breadth will matter as much as technical policy arguments.
The trend: This is one instance of AI governance moving from abstract risk debates to conflicts over whether safety mandates reshape who can build, release, and study advanced models.