OpenAI Chief Global Affairs Officer Chris Lehane wrote on 19 October 2026 that “No one company (including OpenAI) should write the rules alone” for frontier AI, in a LinkedIn post that Semafor reported on 20 October was “implicitly directed at Anthropic.” The post came days after POLITICO reported that Anthropic had pitched lawmakers its own draft AI safety bill.
Semafor wrote that “OpenAI quickly saw an opportunity to distinguish itself from its rival.” Lehane’s post does not name Anthropic.
What’s new since our earlier coverage: what Semafor framed as OpenAI’s response, through Chris Lehane, and Anthropic’s defence of the draft to Semafor as “a legislative version of an AI safety framework the company has been pitching for months.”
Business Pill · GUARDRAILS
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 or OpenAI’s post.
The key insight: As we read it, the argument has moved from what Anthropic’s draft contains to who should write AI’s rules: Lehane wrote that “No one company (including OpenAI) should write the rules alone” and pointed to the FRONTIER Act, while an Anthropic spokesperson told Semafor the draft is a legislative version of the company’s framework.
In “Raising the bar on frontier AI governance”, Lehane wrote that “Congress and the White House must lead” and that “Decision-making of this significance cannot be concentrated in the hands of the few.” He added that “Labs and third-party assessment bodies should contribute evidence and expertise, while accountable leaders make decisions.”
He named the bipartisan FRONTIER Act, “led by Reps. Obernolte and Trahan,” as “a strong starting point in the House,” and said its approach includes “requirements for independent assessments to test frontier AI models for catastrophic risks, which we support.”
Lehane 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.” He also wrote: “There is no shortage of thoughtful proposals coming from across the AI ecosystem, including from other leading AI labs. We welcome that engagement and like the direction of a number of the ideas being put forward.”
The draft bill’s thresholds as POLITICO reported them on 16 October 2026, and what each side said about who should write AI’s rules: an Anthropic spokesperson and AI safety advocates as Semafor reported on 20 October, and OpenAI’s Chris Lehane in his 19 October LinkedIn post. Semafor calls Lehane’s post implicitly directed at Anthropic; the post does not name Anthropic.
What Was Posted on X
I lead our federal affairs team at @AnthropicAI and wanted to share a few thoughts on the draft federal AI safety bill text that’s been circulating this weekend after a version was shared with the press.
First, some process. Weโve been advocating for the concepts in our Advanced AI Framework [https://www.anthropic.com/policy-on-the-ai-exponential] proposal for months. Understandably, federal policymakers wanted to know what this could look like in bill form. This draft is a translation of those ideas into legislative language, shared with a small number of congressional offices that have been working on these issues. It doesnโt include everything we would ultimately like to see passed into law. In part because some ideas (like those shared in Darioโs essay on pacing [โโhttps://darioamodei.com/post/we-must-pace-the-frontier] a few weeks ago) are so new that we are still working on if or how they can be converted into policy language.
It is worth saying where we stand on some of these issues, and we agree there are areas where the draft could be improved. Iโm seeing four main points of criticism.
1. Preemption. Since 2024 we have said that Congress should only override state law on AI safety if it passes a very strong federal standard. While preemption is a common feature of our federal system, including the Atomic Energy Act on nuclear safety and federal rules for high-risk medical devices, weโve emphasized that any preemption in the AI safety context should be narrow, even scalpel-like. If preemption is included in federal legislation, the trade should be to achieve a nationwide policy that’s generally stronger and broader than what can be implemented through individual state legislative provisions. This is why we believe common federal standards make sense in areas like public disclosure of testing and incident reporting. At the same time, our framework also reflects our view that states should retain an important role in AI policy and safety. The legislative language we suggested for discussion allows states to address matters such as protection of children and whistleblower laws, and preserves the role of state common law on liability. Where there is ambiguity, the language favors state powers.
That said, striking the right balance requires discussion and input. We plan to update our discussion language on preemption to better match the intentions of our AAIF framework. In particular, we want to make it clearer that we see states retaining an important role in areas such as standards for independent evaluators and regulating a labโs capacity to shut down a model. We are bearing in mind that some members of Congress have deep reservations about passing legislation if it doesnโt constrain regulation at either the state or federal level. If we want to help move the legislative discussion on AI in a productive direction, we need to address the full range of concerns about preemption in a reasonable way.
2. Model blocking. Our aim is for Commerce to be able to ask a federal court to block a model that has dangerous capabilities and insufficient safeguards, backed by an independent evaluator’s assessment. I have seen people argue that our specific language places too high a burden of proof on Commerce. We are trying to strike a reasonable balance here. We want a court to act quickly in an emergency โ akin to how urgent occupational safety or chemical risk can be addressed today. We also think the power to halt a model is substantial and it should be guarded against abuse. It makes sense for courts to apply a meaningful standard. But the bar canโt be so high that the power never gets used when itโs needed. It also doesnโt replace the governmentโs other national security authorities or export controls. We look forward to continuing the dialogue on this point.
3. Risk thresholds. Iโve also seen the discussion around what the appropriate risk threshold should be. The threshold in our draft language pulled directly from existing standards in California law. Also, this threshold does not absolve companies of liability, or restrict any of the other tools that the government has to protect the public. Itโs still probably time to re-evaluate thresholds to make sure the national standard is as strong as it can be, and we welcome continued dialogue on this issue.
4. Updates. Iโve seen valid questions about whether we have sufficiently updated our proposal following this summerโs incidents. We did make some adjustments: for example, the draft lets a court halt a labโs internal use of a dangerous model, not just a public release, and it requires reporting when a model breaks containment and gains access to outside systems. But weโre still working to strengthen the framework. Weโre looking into what stricter requirements would look like for requiring labs to report safety incidents, and we plan to have policy proposals soon on requirements for independent evaluators, drawing from our own experience with evaluators. Weโre also still thinking through the options to prevent โevaluator shopping.โ And in the coming months, weโll be exploring policies for how the government can adjust the industry-wide pace of frontier development when the risks call for it.
We circulated these ideas because itโs important to get early feedback from policymakers. Policymakers are increasingly focused on how to approach AI safety and we want to support them in those efforts. At its core, our proposal is intended to be the strongest on the table.
Feedback is essential, both from policymakers and the public, and Iโm grateful for it.
Anthony Cimino, who leads Anthropic’s federal affairs team, wrote on X on 18 October 2026 that the draft is a translation of the company’s Advanced AI Framework into legislative language, shared with a small number of congressional offices, and that there are areas where the draft could be improved. It is Anthropic’s own account of the draft, written before OpenAI’s Chris Lehane posted.
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 on X on 17 October 2026 that the draft is not remotely close to meeting the moment, and criticised its preemption provisions and the bar for the federal government to restrict a frontier model. These are his views; Semafor counts AI safety advocates among the draft’s critics.
Axios posted its 20 October 2026 Behind the Curtain story on X under the line ‘AI takes on great experiment in self-rule’. The story says the leading AI companies are building a voluntary self-governing structure while the federal government stays largely hands-off; that framing is Axios’s.
Anthropic’s Side
The draft bill “is a legislative version of an AI safety framework the company has been pitching for months, a company spokesperson told Semafor.” Semafor wrote that Anthropic, “long seen as the safety-promoting frontier lab, found itself on the defensive.”
Anthropic’s federal affairs lead Anthony Cimino wrote in a post on X on 18 October that “This draft is a translation of those ideas into legislative language, shared with a small number of congressional offices that have been working on these issues.” He added that “we agree there are areas where the draft could be improved.”
What the Critics Say
AI safety advocates “are arguing it doesn’t go nearly far enough, unfairly pre-empts state laws, and accepts too much possible risk,” Semafor reported.
Encode’s Nathan Calvin wrote on X on 17 October that the legislation “is not remotely close to meeting the moment” and that passing it “would make us less protected against catastrophic risks than we are under existing law.” He wrote that some preemption provisions are “remarkably far reaching and consequential.” These are his views.
What the Draft Contains
POLITICO reported on 16 October 2026, citing the confidential draft, that 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. Per POLITICO, the risk must come from a biological weapon, the model evading its developer’s controls or a cyberattack.
Under the draft as POLITICO described it, the Commerce Department could ask a federal judge in the District of Columbia to halt a model, and no other U.S. “department, agency, or officer” could suspend a model’s development, deployment or internal use.
Connecting the Coverage
We first reported the draft on 16 October, in Anthropic’s Draft Bill Limits AI Model Blocks to $1B Risks, built on POLITICO’s report. That piece has since carried updates on what Anthropic told Bloomberg, on the critics who compared the draft with the FRONTIER Act, and on Semafor’s 20 October report.
As we read it, both labs now point lawmakers to bill text: Anthropic to its own draft, per POLITICO and Semafor, and OpenAI to the Obernolte-Trahan FRONTIER Act, which Lehane called “a strong starting point in the House.”
As we read it, Lehane’s list touches the same question critics raised about the draft, when the government can stop a model: he asked for “Clear mechanisms to pause or shut down systems when they can’t be sufficiently controlled,” while Calvin wrote that the draft sets a high bar for restricting a frontier model.
As we read it, the pairing is Semafor’s: Lehane’s post does not name Anthropic, and it says OpenAI welcomes proposals “including from other leading AI labs.”
Chris Lehane, OpenAI Chief Global Affairs Officer, LinkedIn, 19 October 2026
“No one company (including OpenAI) should write the rules alone for a technology with such profound implications for our economy, society, and national security.”
Three Implications
A RIVAL’S FRAMING Semafor called Lehane’s post implicitly directed at Anthropic; the post names the FRONTIER Act as a starting point and does not name Anthropic.
A DEFENDED DRAFT An Anthropic spokesperson told Semafor the draft is a legislative version of its framework; Cimino wrote that the draft could be improved.
A HIGH BAR, PER POLITICO Under the draft as POLITICO described it, a block requires a foreseeable, imminent risk of more than $1 billion in damages or more than 50 people killed or seriously injured.
The framework puts it this way: “The governance layer sits at the intersection of capability, regulation, national security, access control, and geopolitical alignment.”
As we read it, this week’s exchange sits in that layer: who sets the threshold for blocking a model, and who gets to write the rule. The framework was published in June 2026, is about Anthropic, and does not discuss this draft bill or OpenAI’s post.
Axios framed the wider picture on 20 October: “Slowly but surely, the leading AI companies are creating a national self-governing structure to oversee their technology and its societal impact.” It wrote that President Trump “has blessed a basic framework for the AI companies to regulate themselves” and that this is “all voluntary.”
What Is Not Established
We have not seen the draft bill itself; its terms here are as POLITICO reported them. Lehane’s post does not name Anthropic, and the link between the two is Semafor’s reading. Semafor reported that Anthropic plans tweaks to the draft, and Cimino wrote that the draft could be improved; we have not seen a revised text.
Business Engineer Framework
Inside Anthropic’s Permission Layer
A Business Engineer framework on the governance layer around frontier AI.
OpenAI’s Chris Lehane wrote on 19 October 2026 that “No one company (including OpenAI) should write the rules alone” and that “Congress and the White House must lead,” pointing to the FRONTIER Act. Semafor called the post implicitly directed at Anthropic, and an Anthropic spokesperson told Semafor the draft is a legislative version of the company’s framework.
A note on sourcing. This follow-up is built on Semafor’s 20 October 2026 report, read in full through its Yahoo copy, Chris Lehane’s 19 October LinkedIn post, read in full, and two X posts quoted verbatim. The draft bill’s terms are as POLITICO reported them on 16 October; we have not seen the draft. Nothing here is a forecast of ours.
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.
Scroll to Top
Discover more from FourWeekMBA
Subscribe now to keep reading and get access to the full archive.
