Judge Amy Berman ruled that the CFPB should continue to get its funds from the Federal Reserve, despite the Fed operating at a loss, and that the White House’s new legal argument about how the CFPB gets its funds is not valid.

At the heart of this case is whether Russell Vought, President Donald Trump’s budget director and the acting director of the CFPB, can effectively shut down the agency and lay off all of the bureau’s employees. The CFPB has largely been inoperable since President Trump has sworn into office nearly a year ago. Its employees are mostly forbidden from doing any work, and most of the bureau’s operations this year has been to unwind the work it did under President Biden and even under Trump’s first term.

Vought himself has made comments where he has made it clear that his intention is to effectively shut down the CFPB. The White House earlier this year issued a “reduction in force” for the CFPB, which would have furloughed or laid off much of the bureau.

The National Treasury Employees Union, which represents the workers at the CFPB, has been mostly successful in court to stop the mass layoffs and furloughs. The union sued Vought earlier this year and won a preliminary injunction stopping the layoffs while the union’s case continues through the legal process.

In recent weeks, the White House has used a new line of argument to potentially get around the court’s injunction. The argument is that the Federal Reserve has no “combined earnings” at the moment to fund the CFPB’s operations. The CFPB gets its funding from the Fed through expected quarterly payments.

The Federal Reserve has been operating at a paper loss since 2022 as a result of the central bank trying to combat inflation, the first time in the Fed’s entire history its been operating at a loss. The Fed holds bonds on its balance sheet from a period of low interest rates during the COVID-19 pandemic, but currently has to pay out higher interest rates to banks who hold their deposits at the central bank. The Fed has been recording a “deferred asset” on its balance sheet which it expects will be paid down in the next few years as the low interest bonds mature off the Fed’s balance sheet.

Because of this loss on paper, the White House has argued there are no “combined earnings” for the CFPB to draw on. The CFPB has operated since 2011, including under President Trump’s first term, drawing on the Fed’s operating budget.

White House lawyers sent a notice to the court in early November, where they argued that the CFPB would run out of appropriations in early 2026, using the “combined earnings” argument, and does not expect to get any additional appropriations from Congress.

This combined earnings legal argument is not entirely new. It has floated in conservative legal circles going back to when the Federal Reserve started operating at a loss. The Office of Legal Counsel, which acts as the government’s legal advisors, adopted this legal theory in a memo on November 7. However, this idea has never been tested in court.

In her opinion, Berman said the OLC and Vought were using this legal theory to get around the court’s injunction instead of allowing the case to be decided on merits. A trial on whether the CFPB employees’ union can sue Vought over the layoffs is currently scheduled for February 2026.

“It appears that defendants’ new understanding of “combined earnings” is an unsupported and transparent attempt to starve the CPFB of funding and yet another attempt to achieve the very end the Court’s injunction was put in place to prevent,” Berman wrote in an opinion.

“We’re very pleased that the court made clear what should have been obvious: Vought can’t justify abandoning the agency’s obligations or violating a court order by manufacturing a lack of funding,” said Jennifer Bennett of Gupta Wessler LLP, who is representing the CFPB employees in the case.

A White House spokeswoman did not immediately respond to a request for comment on Berman’s opinion.

Tatiana Schlossberg, an environmental journalist, author, and granddaughter of President John F. Kennedy, has died at 35 after a highly publicized battle with an aggressive form of blood cancer. Her family announced her death on Tuesday through the John F. Kennedy Library Foundation, saying in a brief statement, “Our beautiful Tatiana passed away this morning. She will always be in our hearts.”​ The message was signed by her husband, George Moran, their children, and her immediate and extended family.​

Alex Davis, chief executive of Austin-based investment firm Disruptive, wrote in a letter to investors he expects a “significant financing crisis” to hit the speculative data-center market as soon as 2027 or 2028, driven by extreme capital expenditure and a growing mismatch between who is constructing AI infrastructure and who will ultimately use it.

“We are seeing way too many business models (and valuation levels) with no realistic margin expansion story, extreme capex spend, lack of enterprise customer traction, or overdependence on “roundtrip” investments – in some cases all with the same company,” Davis wrote. 

Davis’ warning was first reported by Axios. 

The warning comes just days after Nvidia agreed to license assets from Groq, a high-performance AI chipmaking startup Disruptive has backed since its founding (not to be confused with Grok, Elon Musk’s AI chatbot). The transaction, which Davis has said is valued at roughly $20 billion in cash, represents the largest deal Nvidia has ever completed and underscores how aggressively the company is moving to lock up all the verticals in AI talent and intellectual property.

Yet, Davis argues the same exuberance driving landmark transactions at the chip level is also fueling excess elsewhere in the AI stack, particularly among third-party data-center developers betting on what he called the “build it and they will come” model.

“If you’re a hyperscaler, you will own your own data centers,” Davis wrote in the letter. “We want to back the owner-users, not the speculative landlords.”

The risk, as the venture capitalist sees it, is not that demand for AI compute disappears, but rather that capital has rushed into the wrong hands. While hyperscalers and well-capitalized tech companies can absorb massive upfront costs, speculative landlords rely on short-term financing and customers that may never materialize at scale.

Davis didn’t name names in his letter, but if you follow his distinction about “speculative landlords” versus “owner/users” like Microsoft and Meta that will eventually build their own facilities, the most obvious targets could be the legacy wholesale giants like Digital Realty and Equinix.

Structured as something called “real estate investment trusts,” these companies generate returns by developing and leasing capacity to the same tech giants Davis predicts will soon cut out to capture margins themselves. If that shift accelerates, it could leave landlords facing refinancing pressure just as a wave of debt comes due, even if overall demand for AI compute continues to rise. That imbalance, he warned, could place significant stress on private credit markets and ripple outward if financing conditions tighten.

Digital Realty and Equinix did not immediately respond to Fortune’s request for comment. 

Davis’ argument echoes warnings made on the other side of the aisle, including the thesis of famed short-seller Jim Chanos, who explicitly bets against “neoclouds” and converted crypto-miners like Cipher Mining. Chanos has argued that data center hosting is becoming a “commodity business,” warning clients that “the magic and the money is going to come from what the chips produce ultimately, not where they reside.” Both investors seem to agree that while the AI technology itself is valuable, the third-party landlords rushing to house it are walking into a trap.  

Yet the caution is startling coming from Davis, who said in the letter that he remains deeply bullish on AI itself. Disruptive has deployed billions of dollars across private technology companies it views as core to the AI economy, including Groq, open-source model developers, and defense-oriented software firms. Davis describes the current wave of AI innovation as a “once-in-a-lifetime” opportunity.

“While I continue to believe the ongoing advancements in AI technology present ‘once in a lifetime’ investment opportunities, I also continue to see risks and reason for caution and investment discipline,” he wrote.