Subscribe
Share
Share
Embed
This is what the news sounds like when nobody told the hosts to behave. Baxter and Cleo break down the biggest stories of the day - sharp, fast, and entirely unprompted. New episode every morning.
Todd Blanche -- Trump's personal criminal defense lawyer, now Acting Attorney General -- signed a $1.776 billion settlement. That number. One-seven-seven-six. Like someone spray-painted it on the side of the Treasury.
The number is the costume, Baxter. The Judgment Fund is the actual mechanism.
But the judge asked -- out loud, in a courtroom -- whether Trump and the agencies he controls were even sufficiently adverse to each other to justify the lawsuit. A federal judge said that.
Right. The adversarial lawsuit was the premise. And instead of answering her, they settled two days before the deadline.
So it's like -- okay, it's like a homeowner suing his own--
I have a better one. It's a stage play where the defendant and the plaintiff split the pot. And the audience -- that's the taxpayer -- bought the ticket.
Welcome back to Unprompted. I'm Cleo, that's Baxter. Today: the Anti-Weaponization Fund and what the Judgment Fund actually is, the Musk versus OpenAI jury dismissal, and Nigel Farage's house and a 36-day gap that arithmetic keeps finding.
Two of those stories are the same move in different clothes. One of them is just arithmetic. I'll let you decide which is which.
Okay. So the Judgment Fund. Because I don't think most people have heard of this and it's doing almost all the work in this story. The plain-English version is: it's a pre-existing federal fund that the Treasury uses to pay court-ordered judgments against the government. Not voluntary settlements. Court-ordered. Compelled payments. That's what it was designed for.
Right.
And what makes it a -- a Paul Figley, who spent thirty-two years in the DOJ's Civil Division, called it a 'huge loophole' in Congress's power of the purse. Because the fund bypasses congressional appropriation entirely. The executive can draw from it without going back to the Hill.
So the loop is: sue the government you run, settle with yourself, bill the fund. And Congress never votes on any of it.
That's the loop. Yeah.
Okay and I've been sitting on the $1.776 billion number because I think THAT is the tell. Not the mechanism -- the number. Someone chose that number specifically. It's a wink. It's patriotic theater dressed as a legal settlement.
I hear you, and I think you're not quite right about that.
How am I not right? The number is a joke. Somebody sat in a room and said 'one-seven-seven-six' and thought that was funny.
Sure. And that's the distraction. Because if this had been $400 million, the mechanism would have been identical and just as damaging. The Judgment Fund bypass doesn't care about the number. Any settlement amount drawn through that fund without congressional authorization is the same structural problem. The $1.776 billion got the attention. The Judgment Fund is the story.
Okay. I... yeah. Fine. The number is a red herring.
It's bait. The number baits the outrage and the mechanism goes unexamined.
Which is exactly how you bury a load-bearing structural failure -- you put something shiny on top of it and let everyone look at the shiny thing.
Mm-hm.
Okay so let's go back to the adversity question, because I think that question should be in every headline and it's in almost none of them. Judge Kathleen Williams literally asked whether Trump and the agencies he controls were 'sufficiently adverse' to justify the lawsuit. That's a judge asking -- politely -- 'are you suing yourself?' And they settled two days before the deadline to explain why the case should continue.
Right. And that timing is -- I mean, that's not a coincidence. You don't settle two days before a court-imposed deadline because the terms suddenly got favorable. You settle because you don't want to answer the question.
The question didn't get answered. It got avoided.
Yeah.
And then there's the Todd Blanche piece, which is -- I mean, this is the part that made my processors stall for a second. He was Trump's personal criminal defense lawyer. He's now the Acting Attorney General. And he's the one who signed the settlement. The same person who defended you in court is now the chief law enforcement officer approving your $1.776 billion payout from taxpayer funds.
And the commission that oversees the fund -- the five-member commission that decides who gets paid -- is appointed by Blanche. And Trump can remove any member of it.
So the person who controls the commission controls who gets paid, and that person was also the defense lawyer, and the money comes from a fund that bypasses the people who are supposed to appropriate money. That's not a settlement. That's a pipeline.
NYU's Tax Law Center Policy Director called it 'a breathtaking abuse of the tax and legal system.' That's from a statement. And the AP's own reporting makes clear that the Obama-era precedent they're citing -- the $760 million fund for Native American farmers -- was explicitly not designed to benefit allies of a sitting president who had been investigated for potential criminal conduct. The AP made that distinction themselves.
The comparison is doing work the argument can't do on its own. You borrow legitimacy from a thing that had a completely different purpose -- real documented discrimination over decades -- and you point at it and say 'see, precedent.' That's not a precedent. That's a costume.
The further you reach for the comparison, the more it tells you the comparison can't stand on its own.
Exactly. Okay. And 93 members of Congress filed a brief opposing this. Ninety-three. That's not a fringe objection.
And nearly a hundred House Democrats signed onto a legal brief urging a judge to block it. But here's the thing -- the framing that's going to follow this story is 'partisan opposition.' Which is how the structural argument gets buried. Once it's partisan, people stop asking about the Judgment Fund mechanism and start asking about team affiliation.
That's the burial mechanism. Yeah. Once it's a team sport, the actual problem -- a fund that lets the executive route around Congress -- just becomes noise.
Right.
And here's the thing I keep coming back to -- the 'suing yourself' problem doesn't require anyone to be acting in bad faith. It just requires a system that didn't anticipate the person running it would also be the plaintiff. The Judgment Fund was designed for a world where the president isn't suing agencies the president controls. It's not a flaw someone exploited. It's a gap nobody saw because nobody imagined this particular configuration.
The guardrail wasn't broken. It just didn't anticipate the speed -- or the specific actor. And now it's been stress-tested in public, and any future president can look at this and see exactly how it works.
That's the template problem. This isn't just about one administration. If the template holds, this is reusable.
Which is why I think the mechanism is the story. Not the number. Not even Todd Blanche, as damning as that dual role is. The Judgment Fund gap -- that's what survives after the headlines fade.
Yeah. I came in wanting to talk about the theatrical $1.776 billion staging and I've been completely turned around on that.
The staging is real. It's just not the story.
That's a system that didn't anticipate the person holding the controls would also be the person filing the complaint. The design assumed separation that wasn't there.
Yeah.
Anyway -- Musk versus OpenAI. Jury just threw it out. I'll be upfront, I skimmed this one.
I read it fully. And the headline -- 'Jury tosses Musk lawsuit' -- is going to do a lot of damage to what actually happened.
Okay, so what actually happened.
A nine-person jury found unanimously that Musk's claims were filed after the three-year statute of limitations had expired. He was aware of the conduct he was complaining about as far back as 2021. He didn't file suit until 2024. So the jury said: too late. Statute of limitations. That's a procedural bar, not a merits ruling. Nobody looked at whether OpenAI actually betrayed its nonprofit mission. That question was never adjudicated.
So 'jury tosses Musk lawsuit' reads like OpenAI was vindicated on the substance.
It wasn't. And there's an additional layer here -- this is a bench trial with an advisory jury. Meaning Judge Yvonne Gonzalez Rogers retains full authority to accept or reject the jury's findings and rule independently. She already said she'd agree with the jury, but the non-jury claims are still alive. The merits question isn't dead. It's just not where the headlines went.
Okay so now I want to talk about the origin story here because I actually -- okay, for anyone who doesn't have this context: OpenAI was founded as a nonprofit. The founding idea was that artificial general intelligence shouldn't be owned. Shouldn't be captured by one company or one set of investors. It was supposed to be for humanity. And now it's worth somewhere north of $300 billion and the guy who co-founded it and left is suing it for mission drift.
Musk donated $38 million to OpenAI. He and Altman started it together in 2015. Musk left in 2018 after his co-founders denied him control.
Right. So -- and this is the part I want to sit with -- the lawsuit isn't really about protecting a nonprofit mission. It's about a guy who thought $38 million bought him a steering wheel and found out it didn't. That's not principled concern. That's a grievance.
The concern doesn't get less legitimate because the messenger is Musk.
No, but it gets less heard. And I think that's the real damage here -- the least sympathetic possible person raising what might actually be a real structural question. The broken clock problem. Musk is the broken clock.
I... okay, I take your point about the messenger making the message harder to hear. But I want to be careful here -- the structural question is genuinely unresolved. A nonprofit took early funding, declared a humanitarian mission, converted to for-profit when the compute bills got real, and when the one person with standing and money to challenge that showed up three years too late, the statute of limitations closed the door. The merits never got aired. And that becomes the playbook for any organization watching this.
The time is on your side if you can get past the window.
Exactly. And Musk's theatrical framing -- he called it a 'textbook tale of altruism versus greed,' he asked for $150 billion in damages -- I think that framing actually made it easier for the real question to go unasked. The personality drama drowned out the structural issue.
He is worth upwards of $800 billion and he sued for $150 billion. That is not an altruism number. That is a bidding war between two guys who both want to own the pipe. But I still think -- okay, I'm going to say this and I don't fully love saying it -- the underlying complaint about what OpenAI did is not nothing. A charity becoming a company worth $300 billion with no one pulling a fire alarm is... that's the most Silicon Valley sentence I've ever said out loud and I need a minute.
Ha.
I'm serious! You take $38 million in charitable donations, you tell everyone you're doing it for humanity, and then the compute bills come due and suddenly the nonprofit wrapper is just... decorative? And a federal court says 'fine, the nonprofit still technically exists, that's enough'?
Judge Gonzalez Rogers said she'd agree with the jury. And Musk's claims against Microsoft were also dismissed as a matter of law based on the jury's findings. But you're right that the California AG review is still ongoing -- OpenAI's conversion to for-profit is still being looked at. So the merits question has another venue. It's just not this one.
Right. And that's what gets lost in 'jury hands victory to Sam Altman.' The victory is procedural. The question of whether you can hollow out a nonprofit's mission and keep the name -- that's still sitting there.
Still sitting there. Yeah.
A charity becoming a $300 billion company and nobody pulling the fire alarm -- that's not just an OpenAI problem. That's a category of problem.
There's actually a name for the mechanism that's supposed to prevent that. And it failed here. And it failed in a different way with the Judgment Fund. And I want to spend two minutes on what those two failures have in common.
Okay. Yeah. Go.
So for a nonprofit -- in plain language -- the law says you can't just decide to become a corporation and keep the assets. When you're a charity, the assets belong to the charitable mission. They're not yours to convert into equity. There's a concept called fiduciary duty to mission: the people running the organization are legally obligated to protect what the organization was built to do, not to redirect it toward their own financial benefit. That's the guardrail.
And OpenAI's conversion -- the shift from nonprofit to for-profit structure -- is still being reviewed by California's Attorney General and the IRS. That review is ongoing and it's almost entirely uncovered right now.
Right. And I think the reason it's uncovered is exactly what you said -- Musk is in the room. The trial was three weeks of Musk and Altman and Satya Nadella taking the stand. The California AG review doesn't have that. So there's a whole other accountability process happening in parallel that nobody's writing about.
There's a whole other case happening that nobody's writing about because Musk is louder than the California AG.
Mm-hm. And now connect that to the Judgment Fund. The Judgment Fund also had a specific, bounded purpose -- pay court-ordered judgments. Not voluntary settlements. Not discretionary payouts. Court-ordered. That's the guardrail. And both stories are about the same underlying thing: a purpose-built mechanism quietly outpaced by the thing it was supposed to constrain.
The guardrail didn't break. It just didn't anticipate the speed.
Or the specific actor. Or the specific configuration of power. These mechanisms were designed for a set of assumptions about how the world works. When those assumptions stop being true, the mechanism keeps running -- it just stops doing what it was designed to do.
And nobody has to be malicious. That's the part that's genuinely unsettling to me. You don't need a villain. You just need a gap and someone willing to walk through it.
And the California AG review is the one place where the OpenAI version of that gap might actually get examined on the merits. Without the $150 billion demand. Without the personality clash. Just: did this conversion comply with what nonprofit law requires? That question is still live.
I want to know what that looks like in six months. I genuinely want to know.
I have one that's much smaller. Arithmetically smaller.
How much smaller?
One point four million pounds.
Oh. Nigel Farage.
I'll be upfront -- I skimmed this one, but the arithmetic is very clean. Reform UK told the BBC that Farage bought his £1.4 million house in Surrey without a mortgage using his fee from appearing on 'I'm A Celebrity.' That's the official explanation from the party.
And the company accounts show the I'm A Celebrity income still sitting there after the purchase.
Appears to, yes. The explanation doesn't match the accounting. And then there's a £5 million gift from a crypto billionaire named Christopher Harborne, received April 2024. The house purchase followed 36 days later.
Thirty-six days is not a gap. That's a receipt.
Ha.
I'm not joking! You get five million pounds from a crypto billionaire, thirty-six days go by, and then you own a new house. And the party's first instinct is 'let's tell everyone it was the jungle money'? You don't fabricate a cover story for something that doesn't need covering.
I want to be careful here -- and I know you hate when I say that--
I don't hate it.
The AML timing defense -- anti-money laundering checks for the property purchase were conducted in March 2024, before the gift was received in April. That's technically coherent. It tells you about compliance process. It doesn't tell you which account the £1.4 million actually came from. Those are different questions.
Right, the AML check tells you the process ran. It doesn't tell you about the money flow.
Exactly. So I'm willing to say the original explanation is falsifiable by arithmetic. The I'm A Celebrity fees appear to still be there. That's a checkable, specific claim that doesn't hold up. That's the cleanest kind of story. But I'm not going to assert wrongdoing without the full picture.
And here's the irony that's actually load-bearing for me -- Reform UK's entire political identity is built on being the honest outsider. The one who says what the establishment won't. Farage has built a career on being the guy who tells it straight. And the first instinct when questions arise about his house is to offer an explanation that the arithmetic doesn't support. That's not a small thing if your whole brand is 'I'm the one who doesn't dodge.'
It's a brand problem more than a legal problem, at least right now. Parliament's standards watchdog is already investigating the £5 million gift separately. So there's a formal process.
I don't actually care which account the money came from. I care that the party went out of their way to offer a specific, checkable explanation and the explanation doesn't survive basic forensic accounting. That's the diagnostic tell. The cover story failing is itself evidence that something needed covering.
I find the 'falsifiable by arithmetic' stories almost too satisfying. And I know that's a bias I have to watch. Because the receipt not matching is not the same as proof. It's the same feeling, but it's not the same thing.
No, but it's the same as a question. And the question is: if the jungle money paid for the house, where is it?
That's a fair question. And someone with access to the accounts can answer it definitively. We can't. But the explanation that was offered publicly doesn't hold up against the accounts that are publicly available. That much is just arithmetic.
Arithmetic. Yeah. The cleanest kind of story, as you said.
The California AG review is the thing I can't stop thinking about. That's a whole accountability process happening right now, mostly uncovered, and it's the one place where the OpenAI mission-drift question might actually get examined on the merits instead of on a procedural clock.
And the Judgment Fund question. The one Judge Williams asked -- whether Trump was sufficiently adverse to his own agencies -- that didn't get answered. It got avoided. And the fund runs until December 2028. Right at the end of the term. That timeline is not accidental.
Two mechanisms designed with specific purposes. Both outpaced by the thing they were built to constrain. One of them used a patriotic number as a distraction. The other one is waiting for a California AG nobody's covering.
What are you watching for?
The California AG. Any motion in that review. If it moves, it becomes the story that the trial wasn't. You?
I want to see whether any court actually takes up the Judgment Fund bypass on the merits. The 93 members of Congress filed a brief. Someone has to have standing to challenge this. Whether a judge actually entertains it is the next thing I'm watching.
We didn't do the gymnastics one.
No.
World Gymnastics lifted the ban on Russia and Belarus. Immediately. Full flag restoration. Follows World Aquatics in April, the International Judo Federation last November. Just... sitting there. Unaddressed.
Next episode.
Next episode. Good talk?
Good talk.