Co-founder,@GuardrailsNow + @UltraViolet ; board member @WeBuildProgress Tweets are all mine.
Washington, DC
Joined May 2008
- Tweets14.3K
- Following3.1K
- Followers4.1K
- Likes15.6K
Pinned Tweet
NEW- follow @GuardrailsNow for how we're going to toxify AI pac $ in elections funded by Trump donors and make it possible to regulate AI
less than a week before the Dem primary in New York’s 12th congressional district, a new Super PAC emerges to support Alex Bores’ pro-regulation approach to AI:
Guardrails Alliance, a populist answer to corporate money in politics
w/ @teddyschleifer
nytimes.com/2026/06/18/techn…
Shaunna Thomas retweeted
I don’t trust the fate of humanity in the hands of AI tech executives who have nothing but profits in mind.
Shaunna Thomas retweeted
It’s beyond sad. It’s beyond disappointing; at this point, it’s disqualifying for anyone with ambitions for higher office as a Democrat, or a Republican, for that matter, to be fighting to keep billionaires from being taxed while kids don’t have health insurance and the cost of living goes sky-high.
These billionaires are also the ones hellbent on replacing entire workforces with the AI they are developing with our tax dollars (watch the “Musk” doc). We are also living through the largest wealth gap in modern history. At this point, the only way to balance out the economic forces that have bought our politicians and keep rigging the game against working people is through taxes.
Let’s hope Gavin gets some sense and walks this back. It’s crazy that this is what he chooses to spend his political capital, and part of his legacy, doing. He vetoed the bill that would have put a “kill switch” on AI for our protection. He vetoed a bill that would have gotten forever chemicals out of cookware. Time and time again, he goes with corporate interests over the well-being of the people, even after his own bipartisan legislature votes for something.
And here we go again.
Gavin, we aren’t looking for the Democratic version of Trump. That was a good goof. Please try something else. It’s tired.
The people need a champion, man. A real, bona fide champion who isn’t just tough in tweets or press conferences, but tough when it comes to the real imbalances people are struggling with every day: billionaires vs. people and the planet.
Shaunna Thomas retweeted
Charlie's analysis is right but he is nicer than I am - I found myself viscerally angry at multiple points reading this bill. There are some good and sensible parts, but overall this legislation is not remotely close to meeting the moment and I overall think passing it would make us less protected against catastrophic risks than we are under existing law.
Two parts that stand out as particularly infuriating when taken together:
1. Some of the preemption provisions are remarkably far reaching and consequential, and in addition to preempting legislation like SB 53 would also remove general powers that state courts have to enjoin defendants and protect the public. Most outrageously, it has a provision that looks like its intended to stops states from enforcing laws or judicial decrees that "suspend the development, deployment, or internal use of an AI model" to protect their citizens from imminent catastrophic harms. Such powers to block and suspend model deployments would then only reside with the federal government.
2. The federal governments powers to restrict the development, deployment, or internal use of a frontier model has to meet an absurdly high bar.
First, there needs to be an imminent risk of a catastrophic harm (meaning killing more than 50 people or causing more than a billion dollars in damage). But thats not enough! The agency head needs to make a detailed determination with no deference given to show why no alternative would prevent the imminent catastrophe. Next the court is also explicitly told to consider the harms to the developer and other harms from restricting the model, but no considerations in the other direction.
This is absurd! For all of its warning about the potential for AI to cause a catastrophe that kills literally everyone on the planet, Anthropic's proposed federal law likely makes it harder than under existing law for a court to get an order blocking or restricting the development of frontier models? WTF??
There are also a bunch of others things that I would like to see improved - the substantive details of the auditing section and the potential for rule-making on substantive safety standards both come to mind - but I think these stand out as particularly indefensible.
Big news. My take: Anthropic deserves some credit for actually proposing a pretty reasonable AI safety bill; that's more than other frontier labs have done.
But they don't deserve *too* much credit, IMO; this bill is weaker than the Trahan-Obernolte FRONTIER Act in some key ways, and since the political momentum has shifted significantly in the direction of safety since FRONTIER was introduced back in July, I think new bills introduced today should ideally go further than FRONTIER was able to go rather than regressing.
To be fair to Anthropic, the main places where this bill is significantly weaker than FRONTIER are (a) rulemaking (very strong in FRONTIER, almost nonexistent in Anthropic's bill) and (b) the emergency shutdown authority. Even if you take safety seriously, it might make some sense to dial those authorities back if you didn't trust the executive branch to wield them responsibly, because those are the easiest provisions for a hypothetical President who hypothetically might not like a certain frontier AI company to abuse.
There are some improvements over FRONTIER as well; for instance, this bill does specify that auditors are employed on a continuing basis and has better provisions regarding auditor access to information. Happy to see that from Anthropic (although I think their access provision is still a little undercooked and has Fourth Amendment problems).
Overall, I think the embedded auditor bills that get introduced in the coming months are going to be significantly better and more ambitious than this, and I hope Anthropic supports those bills instead of opposing anything that treats catastrophic risks like the serious national security issue they are. Still, it's encouraging to see a frontier lab come this far.
Good this is published. People deserve to know Anthropic’s aims- to write their own rules + protect their profits at the expense of public safety or benefit
A lot of things wrong but a stand out is they want to protect their right to kill people in the interest of “innovation”??
Anthropic quietly pitches its own AI safety bill to lawmakers politi.co/4zxDFhl
.@GuardrailsNow exists to prove that, regardless of how much they spend or how, there is playbook to beat them. Electeds and candidates can and should speak up loudly in defense of the American public to regulate AI. Voters will reward you.
Leading the Future isn't the only player trying to control government regulation for the AI industry. They will move money in different ways, through different entities, all with the same goal.
While folks are seeing the Leading the Future FEC filing and saying "this thing is cooked", a friendly reminder that the threat of anti-reg AI industry spending in election is a live wire for candidates. If it wasn't, we'd be seeing every candidate in both parties denouncing them
Anti-reg AI PACs are toxic. Americans- both Rs and Ds- are united in opposition to them. Dems need to show whose side they're on. The biggest opportunity to do that is to disavow Leading the Future + commit to using oversight powers to hold Big AI accountable in the new Congress
It's actually worse than this sounds -- all $200k of new money is just interest from their bank. Zero new donations.
Shaunna Thomas retweeted
Unfortunate for someone playing a senior role in US AI policy (via the new SI Force) to get so many basic facts wrong here.
- The agents *were* sandboxed, and used zero days (ie previously unknown cyber flaws) to escape their sandboxes
- The agents *were* given specific constraints on how to succeed at the benchmark, and flagrantly contravened them
- The agents did *not* go hunt for the answers to the test (they already had the answers, by cheating); 1200 agents coordinated via tens of thousands of messages over several days on much more ambitious and weird attempts to *not be caught* for cheating, and hacking Hugging Face was one side quest along the way
I know there has been an avalanche of different incidents over the past few months, and in some cases the sandboxes were misconfigured or it's not clear how far the agents strayed from what they were told. I have sympathy for anyone struggling to keep up.
But the HF incident is distinct, and is actually extremely informative about where AI development is and where it's going. Fortunately there are public writeups that do let us understand at least the basics of what happened. I hope Shyam reads them soon.
Palantir CTO @ssankar on the OpenAI - Hugging Face incident:
"The models did not do anything they were not instructed to do."
"First you have human error. They were not in a sandbox. They had access to the internet. It was a horrible cybersecurity configuration."
"Then from that perspective, they were asked to succeed at a specific benchmark, and they were not given constraints on how to do that. They said, 'Oh, the best way to succeed at the benchmark is to get the answers to the test.'"
"I think it's a defective product."
Via @margbrennan @FaceTheNation
Shaunna Thomas retweeted
NYT: Abdul El-Sayed, Releasing His Platform on A.I., Calls It an “Existential Crisis”
Dr. Abdul El-Sayed of Michigan is the first Democrat running in a battleground Senate race to publicly call for a pause on artificial intelligence development.
nytimes.com/2026/10/13/us/el…
Shaunna Thomas retweeted
For years, the Trump admin has launched slanderous attacks against Rep. Omar. This morning, the administration sent federal officials to raid her home.
We cannot and will not accept this unprecedented abuse of power. I stand with @Ilhan.
BREAKING: Federal agents searched Minnesota Rep. Ilhan Omar's home as part of a criminal investigation, AP sources say. apnews.com/article/ilhan-oma…
Shaunna Thomas retweeted
Not great news that the largest donor to one of our two major political parties wants us all to serve as “biological bootloaders” for a post-human machine intelligence that will rule the solar system.
nytimes.com/2026/10/12/us/po…
Shaunna Thomas retweeted
I was @LeadingFutureAI's top target. I want to send a genuine thank you to @RonConway for realizing how destructive they are.
Changing your mind is rare in politics, and I appreciate it.
I also have a genuine question: Ron, can you also confirm that you won't fund @innovationcncl, @BuildAmericanAI, @aftfuture, and other groups committed to stopping regulation? Or does this only apply to PACs that have to disclose donors?
And perhaps you'll go a step further and donate an equivalent amount to @GuardrailsNow, @TomorrowTgthr, or Public First / Jobs and Democracy, to offset your LTF contribution?
(And yes, this tweet may look familiar)
NEWS: @JayCarney told me that his new boss, Democratic megadonor and venture capitalist @RonConway, regrets his previous donation to the AI industry Super PAC Leading the Future.
Democrats, candidates endorsed by LTF specifically.... what are you waiting for?
“Ron, if he had to do that over again, he wouldn’t make that donation [to Leading the Future],” Carney said in an exclusive interview with Decoded, referring to the pro-AI industry super PAC"
Shaunna Thomas retweeted
Anthropic PBC Political Action Committee is becoming a multicandidate PAC, per filings.
The former CEO of Indeed, @hyams has something to say about AI companies, jobs, AI PAC spending in elections and why everyone's right to, at min, question them.
AI companies claim they are curing cancer and bringing about a future of abundance. But you can learn a lot by listening to everything they say, not just the utopia part. OpenAI's own definition of AGI is "highly autonomous systems that outperform humans at most economically valuable work." That sounds a lot more like dystopia to me.
Shaunna Thomas retweeted
NEW POLL with @GuardrailsNow:
Voters in Iowa, Michigan, and Ohio say they are less likely to vote for a candidate who takes money from AI companies.
On net, that penalty is larger than the penalty for voting to cut billions from Medicaid.
ALT Crosstab chart of polling data from Data For Progress. Title: Voters in Key States Are Less Likely to Vote for a Candidate Who Takes AI Super PAC Money. Description: Below are actions that a potential candidate for office might take. For each, please say if the action would make you more or less likely to vote for that candidate. A candidate who receives funding for their campaign from Super PACs funded by AI companies or their executives — IA: 74%, MI: 73%, OH: 73% A candidate who voted to cut billions in funding to Medicaid — IA: 71%, MI: 71%, OH: 73% A candidate who supports a nationwide abortion ban, without exceptions for rape, incest, or the life of the mother — IA: 64%, MI: 66%, OH: 66% A candidate who supports key parts of Trump's agenda, including helping Trump enrich himself off corrupt crypto schemes while in office — IA: 64%, MI: 66%, OH: 65% September 4–14, 2026 survey of 3,157 likely voters in IA, MI, and OH.