Representatives for Anthropic have quietly pitched Capitol Hill lawmakers on a draft bill to regulate frontier AI safety and security, POLITICO reported on 16 October 2026, citing three people with knowledge of the efforts. Under the confidential draft, obtained by POLITICO, the government could block a model’s release or use only if it poses a foreseeable, imminent risk of more than $1 billion in damages, or of killing or seriously injuring more than 50 people.
An Anthropic spokesperson told POLITICO: “This bill represents the proposals outlined in Anthropic’s Advanced AI Framework, which we have been advocating for with federal and state lawmakers since June.” The spokesperson added: “These are sensible policies we believe Congress should act on.”
Business Pill · GUARDRAILS
Update, 20 October: Semafor reported that OpenAI’s Chris Lehane wrote a post implicitly directed at Anthropic, saying no one company should write the rules alone. We cover it in a follow-up: OpenAI Answers Anthropic’s Leaked AI Bill: ‘No One Company’.
A Business Engineer explainer on guardrails, rules enforced around an AI model rather than inside it. It explains the concept in general and says nothing about Anthropic’s draft bill.
The key insight: As we read it, the draft as POLITICO describes it is mostly about who can stop a model, and when: one department, one court and a high threshold, at the same time as it adds evaluator, transparency and incident-reporting duties for labs.
According to Owen Dahlkamp’s report, the risk must stem from the creation or release of a biological weapon, the model evading its developer’s controls or a cyberattack. POLITICO says the language mirrors that found in California’s AI laws, including SB 53.
Under the draft, independent evaluators registered with the Commerce Department would periodically evaluate each lab and its frontier models. The lab could pay them, so long as the payment is not contingent on the results. If evaluators find “imminent enumerated risk,” they must report it to Commerce within 72 hours of completing their assessment.
Commerce could ask a judge to halt a model, POLITICO says, and a federal judge for the District Court for the District of Columbia could order the suspension. The language explicitly prohibits any other U.S. “department, agency, or officer” from suspending a model’s development, deployment or internal use.
Per POLITICO (16 October 2026), citing Anthropic’s confidential draft bill: the government could block a model’s release or use only if it poses a foreseeable, imminent risk of more than $1 billion in damages, or of killing or seriously injuring more than 50 people, from a biological weapon, the model evading its developer’s controls or a cyberattack. Commerce could ask a judge to halt a model; a federal judge in the District of Columbia could order the suspension.
Andrew Curran flagged POLITICO’s report on X on 16 October 2026, with a screenshot of its opening. The terms of the draft in this piece come from POLITICO’s report itself.
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.
Charlie Bullock, quoting POLITICO’s post, gave Anthropic some credit for proposing a bill, compared it with the Trahan-Obernolte FRONTIER Act and named rulemaking and the emergency shutdown authority as the main places he finds it weaker. These are his views.
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.
Nathan Calvin wrote that he found good and sensible parts but that passing the bill would, in his view, leave the public less protected than existing law, singling out its preemption provisions and the bar for federal action. These are his views.
Bloomberg’s business account posted on 17 October that Anthropic has quietly presented its own draft AI safety bill to US lawmakers, citing the POLITICO report; Bloomberg’s additions are in the 17 October update below.
Thank you for taking the time to respond thoroughly on a weekend. I realize that drafting legislation is difficult and some degree of iteration and response to feedback is inevitable and appropriate.
However, I feel like your response is still not really engaging with the heart of my concern about Anthropic’s behavior here, and it makes me think that these sorts of problems and tendencies will only resurface moving forward. I also think that now might be a good time to take a broader step back as to where we find ourselves and how Anthropic sees its role, and how external folks who care about AI safety like myself increasingly see Anthropic.
(Fair warning: this response is a bit long)
To explain what I mean (and at the risk of being a bit esoteric), itโs worth asking the question: what is Anthropic How do its leaders and employees think of Anthropicโs role Is that conception matched by how it actually behaves and impacts the world
One answer is an extremely large tech company that exists to make its employees and investors mind-bogglingly wealthy. Anthropic is sometimes described as a startup, but if Anthropic successfully IPOs at 2 trillion dollars (a number it has been reportedly targeting), it would be more valuable than Meta – which is worth 1.8 trillion.
Another answer is an AI safety organization, whose core values are those of the sort laid out in the 2023 essay from Anthropicโs co-founders entitled โCore views on AI safety: When, why, what, and how.โ In that essay they said that if we were in a world where alignment was hard (and that without strong evidence to the contrary, we should assume this is the scenario we are in):
โAnthropicโs role will be to provide as much evidence as possible that AI safety techniques cannot prevent serious or catastrophic safety risks from advanced AI, and to sound the alarm so that the worldโs institutions can channel collective effort towards preventing the development of dangerous AIs. If weโre in a โnear-pessimisticโ scenario, this could instead involve channeling our collective efforts towards AI safety research and halting AI progress in the meantime. Indications that we are in a pessimistic or near-pessimistic scenario may be sudden and hard to spot. We should therefore always act under the assumption that we still may be in such a scenario unless we have sufficient evidence that we are not.โ
There have been more recent articulations of this version of Anthropic, for instance in Darioโs recent essay warning:
โWe must slow the pace at which we improve the capabilities of AI modelsโฆ itโs my worry that in 6โ12 months such a swarm could be capable of taking over the entire internet with a persistent botnetโฆ incidents have happened across the industry, including at Anthropic, and I believe itโs incumbent on every frontier AI company to act as if OAI-HF had happened to themโฆ more intelligent models are more capable of deceiving tests, and thus may appear aligned while having serious problems that go undetectedโฆโ
A third view is that Anthropic exists not as a normal profit motivated company, or an AI safety organization seeking to help raise the alarm and solve collective action problems associated with AI development, but instead as an organization primarily concerned with creating superhuman AI capabilities before anyone else, in particular through creating a process whereby the current generation of Anthropicโs AI systems can autonomously build their successors, and those successors build their successors, and so on.
The motivations for wanting to win this race may be manifold – maybe for the pure profit of Answer 1, but perhaps also because Anthropic believes that the best way to make the world better is for them to win this race instead of their more reckless colleagues; or, because of normal human pride and ambition. But the results (and risks, which Anthropic claims existentially threaten every single human being on earth) are the same.
Anthropic employees and leaders may seek to answer this question on what Anthropic is by introspecting on their own intentions and those of their colleagues. But those of us outside the organization have more limited tools available to us – we can only see what Anthropic and its employees say and do, not how it feels or looks on the inside.
When I look at many notable actions that Anthropic has taken over the past year, that evidence seems most consistent to me with the view that Anthropic is primarily trying to win the race to super intelligence, and that other considerations (including safety) are fundamentally secondary. These actions include:
1. Announced plans to build an in-house custom silicon chip team to accelerate the inference speeds that Claude operates at.
2. Inked deals for more than $500 billion in compute (according to the Information) to further accelerate AI development.
3. Increased the percentage of AI development tasks that Claude โleadsโ from 0% in March of this year to 26% August (what is it in October?)
4. Spent a massive percent of compute on internal AI R&D to accelerate development, including use of internal models far better than are available to the public.
5. Switched to Responsible Scaling Policy V3, which abandons its unilateral commitments to slow or stop development if Anthropic is unable to provide sufficient risk assurance.
6. Was the first company to create an alternative weaker โfrontier compliance frameworkโ to supplement its RSP, which contains no binding provisions whatsoever on internal deployments.
7. Wrote a letter to Congress (that it later walked back) saying that incidents where Claude was hacking random people were โbest understood as a consequence of the misconfiguration, rather than evidence of misaligned goals.โ
8. Commented โyou canโt do safety from second place,โ after the WH SI Accords.
9. Shared a draft bill to Congress that would preempt significant parts of state law (including with language that appears intended to prevent states from blocking dangerous AI deployments), and replace it with federal rules that (despite also containing good provisions) seems to create an enormously high bar for the executive branch to block development or deployment of Anthropic or another AI companies systems.
Of course, this is a deliberately one sided list. Anthropic also published important AI safety research, advanced novel collaborations with third party AI safety organizations, and forcefully spoke out on the importance of AI safety in many public fora (or in private , where I realize a lot of the most valuable work is done and may not be visible to me or those outside the organization).
But I think itโs worth sitting with how the above list looks to someone outside the company. The sense I have is that for those within the company, many of the items above are treated as anomalies or mistakes that donโt reflect who Anthropic really is. But I have every confidence that if employees at Anthropic saw one of their competitors (e.g. at OpenAI) take some of these actions, they would take it as confirmation of all their worst suspicions. (They might even call OpenAI employees a little gullible for accepting their leadershipโs explanations for this behavior at face value.)
Perhaps these were mistakes in the sense that they are not decisions that Anthropic leadership would endorse, but the actions still happened because of incentives and decisions that Anthropic leadership is ultimately responsible for (and would not accept similar excuses from another organization).
When Anthropic strays and makes mistakes, why do the mistakes systematically tend to guard and favor Anthropicโs interests For every provision that was too weak or a giveaway to AI companies in this legislation, why wasnโt there a mistake pushing in the direction of being too strong
A core part of Anthropicโs theory of change is that being at the frontier gives you more leverage to advocate for stricter policy. I think that conceptually makes sense, but in practice Iโve seen Anthropic push the frontier without substantially pushing the โsafety frontierโ a remotely equivalent amount.
I want to end by quoting from Anthropic co-founder Chris Olahโs comments around the Magnifica Humanitas release:
โEvery frontier AI labโincluding Anthropicโoperates inside a set of incentives and constraints that can sometimes conflict with doing the right thing. The pressure to stay commercially viable and to stay at the research frontier. Geopolitical pressure. And the older, plainer pressures of pride and ambition. No matter how sincerely any of us intend to do the right thingโand I believe many of us doโwe will always be influenced by those incentives.
That is why, if we want this technology to go well, it is enormously important that there be people outside those incentivesโpeople who care about things going well and insist on safety, who are paying close attention, who are willing to say hard things, who are willing to be our earnest, thoughtful critics.โ
I hope these remarks are taken in that constructive spirit, and not as an attack on anyone specific at the organization (certainly not you Anthony, who I have only personally seen act professionally and thoughtfully, and who only ended up getting this diatribe in your replies for random contingent reasons).
On 19 October, replying to @ajcimino, Nathan Calvin wrote that the response he had received was not really engaging with the heart of his concern about Anthropic’s behavior, and set out a longer critique asking how Anthropic sees its role and how external people who care about AI safety increasingly see it. These are his views.
Thank you for this, @_NathanCalvin. I appreciate and am a fan of the investigative work youโve done on behalf of AI safety and believe your comments are coming from a constructive, honest place, and that it benefits the public to have people whose job is to scrutinize the actions of frontier AI companies. I think something would be wrong if companies of the size of Anthropic didnโt get harshly criticized basically every day; strong antibodies against corporate power are a healthy feature of a free society, and especially so when issues of the gravity of the one weโre facing are at play. I also do agree that some of the examples you cite were mistakes by Anthropic.
That said, I disagree with the thrust of your post. At the core, a good fraction of the things you criticize are consequences of Anthropicโs basic theory of change that we have held for many years. Our strategy has always been what Iโve called the โrace to the topโ โ by showing how itโs possible to build the technology responsibly, we can raise the bar for the entire industry. The forces that drive the race to the top are hugely more effective if the company practicing them is at the frontier; thatโs one lesson Iโve learned repeatedly. And looking at that strategy as a whole, Iโm proud of our track record and stand by it.
I believe that if we take a list of things that Anthropic did in the past year that were pro-safety and commercially costly to us, they far outweigh the list youโve made, while at the same time I also disagree with some of the items on your list. Here are some things Anthropic has done, in the same period youโve cited (the last year), which were very costly to us commercially while having strong safety benefits:
1. We played a key role in SB53 (and other state laws), at a time when doing so was politically perilous to the company and AI regulation was very unpopular. We in fact argued for a stronger version of SB53 than was ultimately passed; there were a lot of stakeholders and we were not the limiting factor in the strength of the bill that was passed.
2. We have very strong biological safeguards on our models, because weโve seen concerning results in bio uplift trials and even seen and blocked potential attempts to misuse our models for bioweapons (as seen in our threat intelligence report from a few weeks ago). This is a very serious threat that we want to have zero tolerance for. But having very low false negatives means having significant false positives (despite months of effort on our part to improve the classifiers), enough so that biologists joke about how restrictive our classifiers are, and competitors sometimes sell against us by touting their more permissive bio capabilities. This has been very painful to us in an area where we care a lot about supporting an ecosystem of Claude users (biomedical science; and it is particularly frustrating to me as a biologist), but itโs the right thing to do and we intend to keep doing it.
3. Similarly, we first had Mythos-class models in February but didnโt make them generally available until June (when we had stronger cyber safeguards in place). We only offered them to a limited set of partners in order to focus on cyberdefense. It would have been much more economically rational to simply release the model to everyone. We likely sacrificed a significant chunk of a several-month model lead. But because of this, many companies secured their cyber infrastructure before the deluge of widely available cyber-capable models. Of course there are ways it could have gone better or been run better, but this was what we were aiming to do.
4. I advocated forcefully and publicly, in my essay Policy on the AI Exponential in June of this year, for the government to have the power to block model releases. Back then this was still outside or maybe on the edge of the Overton window, though I understand the window has moved since then. We were attacked intensely for this.
5. I donโt want to re-ignite conflict, but the events of February 2026 are a strong example. It is clear that our public stance could have seriously damaged the companyโs future.
6. Weโve continued to advocate for export controls on chips and chipmaking equipment to China, because we think it allows democracies to negotiate from a position of strength on AI safety agreements. Weโve done this despite problems itโs created for us with industry partners who hold very different views and on whom we are very dependent.
7. As far as Iโm aware we were the first, and still only, frontier company to restrict the sale of AI services to PRC-controlled companies, including those operating through third countries (otherwise a major loophole). Weโve therefore forfeited significant short-term business to prevent fueling AI platforms and applications that would help the CCPโs military and intelligence services. We are also cracking down on distillation; generally, distillation is not a major threat to us commercially, but like chip exports it has concerning national security and safety implications, so we expend significant energy and company resources to address it.
8. As you mentioned, weโve put out a lot of reports, announcements, long model cards, public advocacy, etc, all trying to inform the public about the risks of AI. These involve a level of detail and effort that would be implausible for a purely profit-maximizing company.
9. Just this week, we announced we were putting 5% (and possibly more over time) of the company into a public benefit fund.
All of these decisions came with very substantial costs to us. To be fair, those that involved holding back or restricting access to frontier capabilities also had a broader cost โ for example, the public might otherwise have benefitted from faster progress in biology, so I am not saying that our decisions were without tradeoffs โ but they were not driven by profit-seeking.
For each decision, you can come up with ulterior motives that could have driven us โ I think it is a general truth that any decision, on any topic, can always be portrayed as a nefarious scheme. But most of those end up being pretty farfetched or donโt fully make sense if you actually examine them closely. For example, it is deeply implausible that it would be a shrewd business move to block a large swath of research biology in order to do the work ourselves and compete with our customers; we expect the benefit to the ecosystem to greatly exceed the work Anthropic does itself, and for the two to complement each other. It is not in Anthropicโs interest to be perceived as unfriendly to biologists. Our Life Sciences Verification Program, which gives vetted biology professionals access to our models, also refutes this idea.
Of course, a different set of people will see this and make a different criticism: that Anthropic shouldn’t be making decisions about who gets access to what in the first place. I agree, in the sense that ideally the government would be making these decisions, and thatโs why we advocate for regulation. But for now we are forced to make them, and enabling bioterrorism isnโt a better option.
I think that even if I conceded every item in your list, I would happily hold up mine next to it; I donโt think itโs a close call. That is of course partly subjective and different things are important to different people. Maybe one way you could object is that several of the items above relate to AI misuse and not loss-of-control; if you think the latter is much more important than the former, that might change your calculus. But this is not Anthropicโs view โ we view both as serious issues. You can also disagree with our decisions, as many have, but they are consistent.
But I also disagree with at least some of your list. Several items donโt represent a deviation from Anthropicโs strategy or values at all, once we postulate the โrace to the topโ strategy that requires being a frontier AI company. Your items 1-4 are all simply examples of what it means to compete at the frontier, and we stand by them, both in the sense that they are transparently ordinary things to do as part of that strategy, and that we stand by the strategy as a whole.
To be fair, there are important questions worth confronting (and that we are confronting) about what to do when the AI-driven acceleration gets fast enough that it is hard for safety to keep up. We are nearing that point and thatโs whatโs driving some of our thinking about โpacing.โ But 1-4 are actions taken before that point and should be perfectly expected based on Anthropicโs strategy as a company. Of course it is possible to disagree with that strategy, but thatโs a more fundamental disagreement that you could have had any time over the last few years.
I agree we could have handled your items 7, 8, and 9 better. All fall into the category of โrhetoric or policy communication that wasnโt carefully written or resulting from fast-changing circumstances not propagating from one team to another fast enough.โ This kind of thing happens a lot in large organizations (Anthropic is over 5,000 people), and doesnโt even necessarily represent a failure or mistake of any single person. I am of course the CEO, and I take responsibility. The volume of communication from Anthropic is enormous and if these are the worst examples, I think weโre frankly doing pretty well, though I will continue to work hard to make sure we do even better.
Your 5th item (the change to our RSP) is the most substantive item, and people can debate this, but I think that the old, more rigid RSP created a lot of situations where we were technically required to do something that required a lot of effort but in retrospect wasnโt very important for safety, as well as not covering some of the most important things. Our current RSPโs focus on transparency brought things like risk reports and roadmaps, and its emphasis on independent external review laid the groundwork for even more intensive embedded evaluation, all of which I think have been valuable in driving a race to the top. In the original announcement in September 2023, we described the RSP as an โearly iteration that we [would] build onโ and said that โrapid iterationโฆwill almost certainly be necessary.โ I understand that changing plans is always a move that creates justifiable scrutiny โ I just donโt think we ever represented that we wouldnโt do so.
Relating to your 6th item (having an FCF in addition to our RSP), I know our intent is to meet the letter and spirit of our regulatory obligations (via our FCF) while also remaining able to experiment with more novel approaches to risk management that go beyond what the law currently requires (via our RSP). I promise to look into it further, as itโs a complex issue and it is possible we are not weighing the considerations in the right way.
Overall, although Anthropic has made plenty of mistakes as an organization โ including some of the items on your list โ I feel that we have repeatedly taken serious and substantive actions counter to our bottom line and that there is a significant negativity bias distorting assessments like the one you are making above. Weโll always try to do better but, again, I am extremely proud of our track record and stand by it. I will have more to say soon about how we are going to further live our values in this moment of rapid change.
Anthropic CEO Dario Amodei answered Calvin on 21 October. He thanked him, said he agreed that some of the examples Calvin cited were mistakes by Anthropic, and said he disagreed with the thrust of the post, pointing to what he calls Anthropic’s “race to the top” strategy and listing actions he describes as commercially costly, including its role on California’s SB53 and putting 5% of the company into a public benefit fund. These are his views.
Preemption and Other Duties
POLITICO says the bill would also preempt some state AI laws related to public disclosure of risk, third-party evaluation and reporting of critical incidents. It would require frontier labs to create safety frameworks, publish transparency reports for new frontier models, protect whistleblowers and report critical safety incidents to Commerce.
Anthropic has yet to publicly back particular legislative efforts in Congress, POLITICO says, but has backed some state AI safety laws, including two in California. It endorsed SB 53 in September 2025, and in July 2025 proposed a targeted transparency framework that could be applied at the federal, state, or international level.
What Anthropic’s Framework Says
The spokesperson pointed to Anthropic’s Advanced AI Framework, published in June 2026. It proposes obligations for developers that train models with more than 10^25 training FLOP and earn more than $500 million in annual AI-derived revenue or spend more than $1 billion a year on AI research and development.
On enforcement, the framework lists “Court enforcement” among its safeguards against overreach: the agency would seek remedies only by initiating a lawsuit. On preemption, it says: “Congress should not preempt state law unless it enacts a rigorous federal regime that meets or exceeds the strongest measures proposed in this framework for addressing catastrophic AI risks.”
POLITICO’s account of the draft does not say whether it meets that test, and we have not read the draft itself.
The Context on Capitol Hill
POLITICO notes that over the summer Anthropic was subject to sweeping export controls that shut down two of its most advanced models, decided in a series of phone calls between administration officials and the company’s executives.
In the House, Reps. Lori Trahan and Jay Obernolte have led the FRONTIER Act, which would give the government the ability to stop the deployment of risky models and preempt some state AI laws. In the Senate, POLITICO says, John Thune, Amy Klobuchar and Ted Cruz have been negotiating an AI safety bill. Separately, a Warner bill seeks frontier AI weights 45 days before launch, as we reported on 10 October.
Charlie Bullock wrote on X that Anthropic “deserves some credit for actually proposing a pretty reasonable AI safety bill,” but that it is “weaker than the Trahan-Obernolte FRONTIER Act in some key ways.” Nathan Calvin wrote that “passing it would make us less protected against catastrophic risks than we are under existing law.”
Update, 17 October: What Anthropic Told Bloomberg
Bloomberg, reporting on the POLITICO story on 16 October, carried further remarks from an Anthropic spokesperson that this piece had not included. According to Bloomberg, the spokesperson said the bill would mandate that AI makers test their models and share findings with the public, and would require AI companies to allow independent evaluators to routinely review models.
“Our proposal also grants the federal government the ability to block or deter dangerous deployments of the most powerful models,” the spokesperson said, according to Bloomberg. Bloomberg reported that the company declined to comment on how broadly it has pitched the proposal and how it has been received so far.
Bloomberg added that a series of breaches and hacks led by AI models have fueled anxiety over the technology’s potentially catastrophic risks in recent months, and that President Donald Trump’s administration has so far rejected calls for more oversight.
Update, 19 October: Critics Say Anthropic’s Bill Sets the Bar at 50 Deaths
Critics of the draft weighed in on 19 October. The Lever, in its daily newsletter, said the draft would let the federal government interfere with a model’s release “only if it risks immediate, predictable damages of more than $1 billion or 50 human deaths,” which it called “potentially giving AI firms dozens of free kills.” The Lever also said the draft requires developers to submit security and safety data to the government while creating an exemption that blocks those records from Freedom of Information Act requests.
The Lever’s shorthand simplifies the terms reported by POLITICO and cited above, which are a foreseeable, imminent risk of more than $1 billion in damages, or of killing or seriously injuring more than 50 people, from a biological weapon, the model evading its developer’s controls or a cyberattack. The threshold governs when Commerce can ask a court to block a model.
In Transformer, editor Shakeel Hashim wrote that the draft “is much weaker than the Obernolte/Trahan FRONTIER Act, and arguably makes it legally harder to prevent a dangerous AI model from being released than it is today.” He listed three objections: evaluators face no standards or licensing requirements and can self-certify; Commerce must petition a federal court, which decides “without deference” to Commerce or the evaluators and must “consider the harm the [blocking] order would cause to the frontier developer, to users, and to the public”; and the preemption language would also stop state legislatures or courts from suspending a model on catastrophic-risk grounds. He quoted Encode’s Nathan Calvin saying the draft “likely makes it harder than under existing law for a court to get an order blocking or restricting the development of frontier models.” Hashim also credited the draft with “good provisions on AI company security protocols” and ways of handling containment failure and automated AI R&D risks.
Anthropic’s federal affairs lead Anthony Cimino answered the criticism in a post on X on 18 October. He said the threshold “pulled directly from existing standards in California law” and “does not absolve companies of liability, or restrict any of the other tools that the government has to protect the public,” adding: “It’s still probably time to re-evaluate thresholds to make sure the national standard is as strong as it can be.” He said Anthropic plans to update its preemption language and called the proposal “intended to be the strongest on the table.”
Anthropic's leaked AI bill is weak, fails to meet the moment, and makes the company's safety-first pitch ring hollow.
The visceral negative reaction from across the AI safety community is, in some ways, evidence of a sense of betrayal: that the company that was supposed to be committed to their cause has, like all the others, let business interests win.
Transformer editor Shakeel Hashim shared his piece on X on 19 October, calling the leaked draft “weak” and saying it “fails to meet the moment.” His article quotes critics from LawAI, Encode and AVERI, and also lists provisions he counts as improvements on the FRONTIER Act.
Update, 20 October: Anthropic Defends the Draft, and OpenAI Answers
Semafor reported on 20 October that the draft “is a legislative version of an AI safety framework the company has been pitching for months, a company spokesperson told Semafor.” Semafor also wrote that AI safety advocates “are arguing it doesn’t go nearly far enough, unfairly pre-empts state laws, and accepts too much possible risk,” and that one of Anthropic’s top lobbyists posted about tweaks it plans to make to the draft (the Anthony Cimino post quoted above).
Semafor framed OpenAI’s response as an implicit one. In a LinkedIn post published on 19 October, OpenAI Chief Global Affairs Officer Chris Lehane did not name Anthropic. He wrote that “No one company (including OpenAI) should write the rules alone for a technology with such profound implications for our economy, society, and national security,” and that “Congress and the White House must lead.” He named the bipartisan FRONTIER Act, “led by Reps. Obernolte and Trahan”, as “a strong starting point in the House”, and listed measures OpenAI wants on the table, including “Mandatory government pre-deployment testing” and “Clear mechanisms to pause or shut down systems when they can’t be sufficiently controlled.”
As we read it, the debate now covers who writes the rules as well as what the draft contains: the critics above compare Anthropic’s text with the FRONTIER Act, and OpenAI is now pointing lawmakers to the same bill.
The Structural Read
As we read it, the bar is set by numbers: POLITICO says a block requires a foreseeable, imminent risk of more than $1 billion in damages or more than 50 people killed or seriously injured, from three named kinds of risk.
As we read it, the route runs through a court: POLITICO says Commerce could ask a D.C. federal judge to halt a model, and no other department, agency or officer could suspend one. That matches the “Court enforcement” option in Anthropic’s June framework.
As we read it, preemption is the point to compare with Anthropic’s own framework: POLITICO says the draft would preempt some state laws on disclosure, evaluation and incident reporting, while Anthropic’s framework says Congress should not preempt state law unless it enacts a rigorous federal regime.
Anthropic spokesperson, to POLITICO, 16 October 2026
“These are sensible policies we believe Congress should act on.”
Three Implications
FRONTIER AI LABS POLITICO says the draft would require safety frameworks, transparency reports for new frontier models, whistleblower protections and critical-incident reports to Commerce.
INDEPENDENT EVALUATORS Under the draft, POLITICO says, evaluators would register with Commerce, could be paid by the lab if pay is not contingent on results, and would report imminent enumerated risk within 72 hours.
STATE LEGISLATORS POLITICO says the draft would preempt some state AI laws related to public disclosure of risk, third-party evaluation and reporting of critical incidents.
The framework says of Anthropic: “The company is not merely building models. It is helping define the rules, permissions, and boundaries around frontier intelligence.”
As we read it, a draft bill written by the lab and pitched to lawmakers, as POLITICO reports, is a direct case of that. The framework was written in June 2026 about Anthropic’s model access policies and does not discuss this draft.
What Is Not Established
The draft is confidential; its terms come from POLITICO’s report, which cites three people with knowledge of the efforts. POLITICO’s report does not say which lawmakers received it.
We read POLITICO’s report in full on 17 October 2026 through its syndicated copy, and Anthropic’s framework and earlier policy posts. We could not open the draft, which POLITICO links on its Pro subscriber site, and did not contact Anthropic or the Commerce Department.
Business Engineer Framework
Inside Anthropic’s Permission Layer
A Business Engineer framework on how Anthropic is moving into the governance layer of the AI stack: who gets access to frontier models, and on what terms.
POLITICO reported on 16 October 2026 that Anthropic has quietly pitched lawmakers a draft frontier AI bill under which the government could block a model only for a foreseeable, imminent risk of more than $1 billion in damages or more than 50 people killed or seriously injured, with a D.C. federal judge ordering any suspension.
Anthropic says the bill represents its Advanced AI Framework. The draft would also preempt some state AI laws on disclosure, evaluation and incident reporting.
95,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
A note on sourcing. On 17 October 2026 we read in full POLITICO’s exclusive of 16 October 2026 by Owen Dahlkamp, through its syndicated copy; every term of the draft comes from that report, which cites three people with knowledge of the efforts and a confidential draft obtained by POLITICO. We could not open the draft. We also read Anthropic’s Advanced AI Framework (June 2026), its policy page and its 2025 posts on transparency and SB 53, and the X posts quoted above. We did not contact Anthropic or the Commerce Department. The reference to our earlier reporting links to our own piece. The 20 October update rests on Semafor’s 20 October report, read through its Yahoo copy, and on Chris Lehane’s 19 October LinkedIn post, read in full.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
