Anthropic’s Call for A.I. Nonproliferation
The artificial intelligence giant said a “brake pedal” was needed to protect humanity from self-improving models. The proposal could have big consequences.
You’re reading the DealBook newsletter. The most crucial business and policy news you need to know from Andrew Ross Sorkin and team.
Andrew here. In 2023, Elon Musk signed an open letter, along with over 30,000 other signatories, seeking a pause on artificial intelligence. Now, Anthropic is suggesting a similar break, arguing the risks could outweigh the rewards. All of this as it plans to go public.
Also: The company behind the S&P 500 has declined to change its rules to quickly add SpaceX to its index after the I.P.O., unlike Nasdaq. (A reminder: Nasdaq won the SpaceX listing on its exchange.) We go into all that below.
A global brake pedal for A.I.?
Artificial intelligence giants like Anthropic and OpenAI are racing toward blockbuster I.P.O.s that could value them at more than $1 trillion each, on the promise of their rapidly advancing products.
Yet Anthropic’s new suggestion that A.I. labs should weigh pausing work on their bleeding-edge technology — in the name of safety — raises questions about the risks of investing in these companies. (That could have big repercussions if the federal government takes stakes in them, as the news outlet NOTUS reports.)
From a blog post on Anthropic’s website on Thursday:
We believe it would be good for the world to have the option to slow or temporarily pause frontier AI development to enable societal structures and alignment research to keep up with the advance of the technology.
Jack Clark, an Anthropic founder, told BBC News, “Right now, it’s like the A.I. industry has a gas pedal, but it doesn’t have a brake pedal.”
The reason: “recursive self-improvement.” That refers to how A.I. models could soon be able to improve themselves without human intervention. Increasingly capable models and the rise of agents that can run autonomously would make that possible.
As of last month, Anthropic noted, 80 percent of the code added to the company’s code base was written by its Claude model.
Anthropic is proposing the A.I. equivalent of a nuclear nonproliferation treaty:
A meaningful slowdown or pause would require multiple well-resourced labs at or near the frontier, in multiple countries, agreeing to stop under the same conditions. It would also require that each can verify that the others have actually stopped.
The company added that its Anthropic Institute, an in-house research arm, would work on ways to create such a system.
A.I. oversight is an increasingly important concern. Polls suggest that Americans worry more about building guardrails around the technology than about speeding up its development. Even the Trump administration has become more open to actively regulating A.I. companies.
But pausing A.I. development could have serious consequences:
Anthropic and OpenAI have attained astronomical valuations thanks to their rapid growth. (Remember that Anthropic was recently valued at $900 billion.) Disrupting that could crater their stocks and hammer investors.
The system Anthropic outlined would require effective policing of labs and international cooperation among rival nations like the U.S. and China.
Anthropic’s call drew skepticism, including from David Sacks, the administration’s former A.I. czar who has long criticized the company. “In other words, you want the government to save us from… you,” he wrote on X.
Several critics have argued that Anthropic has made fear-mongering a marketing strategy, though industry experts have told DealBook that company executives appear genuinely concerned.
HERE’S WHAT’S HAPPENING
Blackstone caps withdrawals from the world’s biggest private credit fund. The asset management giant limited redemptions from the $79 billion flagship fund, known as BCRED, to 5 percent of assets per quarter, after investors asked to withdraw double that amount. The firm had avoided the redemption caps instituted by its competitors even as investors worry about the health of private credit firms’ loan books.
The Supreme Court says the S.E.C. can recover illicit gains without proving victim losses. The court ruled that the securities regulator didn’t need to prove that investors had suffered a financial loss in order to strip money from wrongdoers. The case, which involved a trader who pleaded guilty to securities fraud but tried to keep the $4.1 million he’d made, comes as the Trump administration has eased up on financial regulations.
Investors await the latest jobs report for signs of slowing growth. Friday’s data release from the Bureau of Labor Statistics is expected to show that the U.S. economy added 85,000 jobs in May, a three-month low, and unemployment unchanged at 4.3 percent, according to economists surveyed by Bloomberg. Growing layoff announcements and a war-related jump in inflation continue to put pressure on the economy.
Not so fast
SpaceX’s I.P.O. has already captivated Wall Street — with one notable exception.
S&P Dow Jones Indices said on Thursday that it would not change its rules to fast-track the inclusion of a new mega-cap company, like SpaceX, in one of its stock indexes.
The announcement is a blow to Elon Musk’s rockets-and-artificial-intelligence company, as well as to the I.P.O. hopefuls Anthropic and OpenAI.
They will have to wait at least 12 months from their public trading debut to qualify for entry to the S&P 500, a colossus.
The move is likely to rekindle a major debate at the heart of SpaceX’s listing, which is expected to value the company at $1.77 trillion. Should Wall Street firms make exceptions to add newly listed giants to stock indexes within days of going public? (Nasdaq and FTSE Russell did just that.)
The question isn’t just about fairness. Investment funds that track the indexes — and which dominate investors’ portfolios — are often obligated to buy up new index entrants, affecting market pricing.
A reminder: While some A.I. bulls are thrilled by speedy entry, major pension funds don’t want to see a change in the rules.
Either way, brokerages like Robinhood and E*Trade from Morgan Stanley have been bracing for a tsunami of SpaceX orders from retail investors — a group that Musk is again embracing to give the stock liftoff.
(Complicating things, underwriters have reportedly been told by Washington that they cannot take orders for SpaceX shares from investors in China and Hong Kong, according to Bloomberg.)
Why so much buzz? Jamie Dimon of JPMorgan Chase hosted Musk for a livestream event on Thursday at the bank’s New York headquarters. And as the roadshow winds on, more to-the-moon numbers are emerging. Analysts at Goldman Sachs, which is leading SpaceX’s offering, see the company’s revenues rising to $474 billion in 2030, a roughly 25-fold increase over last year, according to The Financial Times, citing an unnamed source.
A.I. is expected to drive the brunt of that growth, The FT added. (That said, Martin Peers of The Information is skeptical: SpaceX’s A.I. revenues largely come from … ad sales at X, which is part of the xAI division, yet Goldman’s projections require xAI to surpass OpenAI by 2030.)
A successful listing could make Musk the world’s first trillionaire. Plenty of others are expected to reap huge returns, too. Om Malik, a tech journalist, details a few:
Peter Thiel’s venture capital firm Founders Fund, which made a $20 million investment in 2008. Its stake could be worth around $182 billion today, Malik calculates.
Alphabet, Google’s parent company, and Fidelity, which co-led a round in 2015 at a $122 billion valuation.
Sequoia, which invested in 2019 and 2021. The second investment was at a $74 billion valuation.
Andreessen Horowitz, which led a $750 million round in 2023 at a $137 billion valuation.
Nvidia and Gulf region investors, including Qatar Investment Authority and the Abu Dhabi-based MGX, which have stakes via SpaceX’s merger with xAI.
A must-read: the updated analysis of SpaceX’s financials by Aswath Damodaran, a finance professor at N.Y.U.’s Stern School of Business and a prominent valuation expert.
“You don’t need to keep asking Code Puppy the exact same question again.”
— Suresh Kumar, the C.T.O. of Walmart, on the retail giant’s capping employees’ use of the company’s internal artificial intelligence coding tool, Code Puppy. (The company realized that some employees had been using it to perform the same tasks repeatedly.) It comes amid growing concerns over workers’ use of expensive A.I. tools and “tokenmaxxing.”
Talking A.I. with the C.E.O. of Banyan Software
Every week, we’re asking leaders how they use artificial intelligence. This week, David Berkal — who leads Banyan Software, a company that acquires and holds enterprise software companies in industries like health care and finance — told Sarah Kessler that he’s not worried about A.I. eating software. The interview has been condensed and edited for clarity.
How do you personally use A.I.?
I’ve built a chief-of-staff agent. It’s connected to my calendar, my email. It has context on all of my meetings. It knows my annual goals. It knows who the key people are in the organization and even my family.
It’s telling me: “Here are the meetings coming up. Here’s what might be important in those meetings.” It will say, “There’s this important internal initiative, and you’re not spending enough time on it,” and proactively put time on my calendar to work on it.
Your company acquires enterprise software firms, a sector that has been hit by investors’ A.I. disruption fears. What are they missing?
Horizontal software companies are, frankly, pretty easy to vibe code. We’ve canceled software subscriptions left, right and center that we can just vibe code ourselves.
The companies we focus on are mission critical and very deeply embedded into their customer workflows. They typically operate in regulated markets. You don’t just vibe code credit union banking software or fire station dispatch software over the weekend.
What advice do you give your portfolio software companies about incorporating A.I.?
The key is getting from what I would call Level 1 A.I., which is just an easier way to query data with a chatbot, to Level 2, which is, you’re really transforming the work, productizing data. You are going from a system of record to a system of action.
THE SPEED READ
Deals
Helion said it had raised $465 million in a round led by Thrive Capital, nearly tripling the valuation of the nuclear fusion start-up, to $15.5 billion. (Reuters)
Bill Ackman’s hedge fund, Pershing Square, ultimately failed in its bids to take over Universal Music Group. But selling its stake in the music label will reportedly net the firm a $600 million profit. (WSJ)
Technology and artificial intelligence
Dara Khosrowshahi, Uber’s C.E.O., said the ride-share company is recruiting fewer new drivers in cities that have autonomous taxis. (Fast Company)
Robinhood is moving some of its customers’ World Cup bets away from Kalshi to Rothera, an exchange it bought a stake in last year. (Bloomberg)
Jane Street Group, a big trading firm, is reportedly planning to build its own data center to deal with shortfalls in computing power, and train its own A.I. models. (Bloomberg)
Best of the rest
“The Wall Street Mania Pushing Knicks Tickets to $176,000” (WSJ)
Why the global flight of rich people planning to move to a tax-friendly destination is slowing. (FT)
We’d like your feedback! Please email thoughts and suggestions to deal...@nytimes.com.
June 5, 2026
:An earlier version of this article misstated the size of a 2023 SpaceX investment round. It was a $750 million round, not $750 billion.
Andrew Ross Sorkin is a columnist and the founder of DealBook, the flagship business and policy newsletter at The Times and an annual conference.
Bernhard Warner is a senior editor for DealBook, a newsletter from The Times, covering business trends, the economy and the markets.
Niko Gallogly is a Times reporter, covering business for the DealBook newsletter.
Brian O'Keefe is the managing editor of DealBook, a newsletter from The New York Times that covers business, policy and culture — and the many ways they overlap.
More from the DealBook Newsletter
DealBook
Our columnist Andrew Ross Sorkin and his Times colleagues help you make sense of major business and policy headlines — and the power-brokers who shape them. Sends daily. Get it in your inbox.