The White House move to restrict access to Anthropic's latest AI model — using what is known as export controls — could harm the long-term financial prospects of the entire U.S. AI industry.

Why it matters: Anthropic and OpenAI's valuations depend on the global adoption of their most advanced models, and government restrictions could limit that growth.

Zoom in: If this move is more than a temporary blip, "it's not great news for U.S. tech firms or for those assuming breakneck speed of AI adoption," Jim Reid, global head of macro at Deutsche Bank, wrote in a research note.

• Hundreds of billions of dollars are being spent by the data center hyperscalers and the AI labs to fund their ambitions, seeking to eventually profit from having the best models.
• Their calculations can work only if the government doesn't cut off access every time they achieve that goal.

How it works: Businesses that pay for AI models need to make sure they can keep access to them.

• "You can't rely on something that could be switched off," Reid says.

The big picture: It's not just about Anthropic.

• "Everyone who uses AI will see the writing on the wall that future AI models from OpenAI and Google are also going to be seen as having potential serious security risks," says Martin Chorzempa, a senior fellow at the Peterson Institute who studies AI and fintechs.

Follow the money: Companies are already wary about locking in contracts with major AI labs in case a competitor comes out with a better model.

• Now, they can add "potential regulation" to the list of reasons to keep their AI tools diversified.
• If companies don't want to sign contracts with OpenAI or Anthropic, that could put a ceiling on revenue growth for the two AI labs just before both are expected to go public later this year.

Yes, but: The models Anthropic can no longer offer were pricey for them to run. AI labs typically subsidize the costs of running their most powerful models in the beginning.

• Anthropic was rolling out these models for only two weeks to paid subscribers, for example, and then users were going to have to pay a usage fee on top of their subscriptions to access them.
• The government in effect shortened Anthropic's subsidy window for its most expensive model ever.

Between the lines: Export controls can be a powerful tool of leverage — recall when China cut off access to rare earths to get the upper hand in trade negotiations with the U.S.

• But they also have significant downside risk: Countries and companies will start looking for alternatives.
• "The challenge with export controls is anytime you do it, you encourage the development of alternative suppliers," Chorzempa says.
• In the case of rare earths, other countries are now looking to mine their own.

Zoom out: Even before the move against Anthropic, there were already concerns, particularly in Europe, over the U.S. government using AI tool access as a lever of geopolitical influence.

• The move could provide some momentum to Chinese AI models that are open source.
• "You have no idea whether the U.S. government is just going to shut off your access to any future models," Chorzempa says. "That's a big advantage to open models."

Databricks is launching new tools to help companies cap AI costs, after finding customers had accidentally spent tens of millions of dollars on their broader AI bills in a single month.

Why it matters: AI agents are making corporate software bills harder to predict, and Databricks wants to be the layer companies use to keep them under control.

The big picture: The rise of AI agents that autonomously interact with models without much human supervision is increasing AI usage and forcing companies to blow budgets.

• Traditional cloud cost tools often flag overspending only after the damage is done.
• Databricks hopes its new tool helps firms pivot from token maxing to "value maxing," Patrick Wendell, Databricks co-founder, tells Axios.

What they're saying: Wendell says he's seen companies go from little or no AI spending to accidentally spending tens of millions of dollars a month.

• "We've definitely seen mistakes that are in the millions," he added.
• Wendell said AI token costs are entering the top three highest expenses among customers behind salaries and other IT expenses. That's a problem the team has been working to address.

Driving the news: The new Databricks product — Unity AI Gateway — will include AI spend limits, protections to prevent "runaway spend" and recommendations designed to help companies manage AI costs across multiple providers.

• The gateway can recommend cheaper models for tasks that don't require the most token-heavy or expensive options.
• It will monitor individual user sessions, which will then inform feedback on efficiency of AI usage, which could involve removing employee access or moving them to a cheaper model if they aren't using the tools efficiently.

Zoom out: Monitoring individual AI use could be a tough sell to employees.

• Wendell argued the pushback is usually stronger when companies use employee activity for AI training, rather than cost control.
• "The amount of data we need for this purpose is relatively narrow, and it only is related directly to their use of these coding tools," Wendell said.

The bottom line: Better cost controls could help companies rein in AI bills — but they could also pressure model providers counting on fast-growing enterprise usage.

Disney is using custom Adobe Firefly AI models trained on its own characters to help Imagineers design concepts and prototypes for theme parks and hotels.

Why it matters: The partnership shows Disney is willing to deploy generative AI when it can control how its intellectual property is used and protected.

Driving the news: Disney says it will use Adobe Firefly Foundry, which combines Adobe's commercially safe AI models with Disney characters from franchises including Frozen, Moana, Lilo & Stitch and Cars.

Keep reading.

• Nvidia's $25 billion bond sale lets the chip giant join the growing list of companies selling bonds to fuel their AI ambitions. (Reuters)

Mady here after reading this New York Magazine profile of a Disney employee whose entire digital life was exposed in a malware attack.

Why it matters: If you're as scared of horror movies as I am, just read this article to get a similar hit of fear without the jump scares.