Minimum Competence

This Day in Legal History: Earl Warren Takes the Center Chair

On October 5, 1953, which was the first Monday in October and the opening day of the Supreme Court’s term, Earl Warren was sworn in as the 14th Chief Justice of the United States. Chief Justice Fred Vinson had died of a heart attack less than a month earlier. To make sure the Court opened its term with a full bench, President Eisenhower gave Warren, then governor of California, a recess appointment. Eisenhower and Vice President Nixon attended the ceremony. The Senate confirmed Warren the following March, and he took the oath again.

The timing mattered. Brown v. Board of Education had already been argued once and was set for reargument that term. Warren spent months building a unanimous Court, and in May 1954 he delivered the opinion holding that racially segregated public schools violate the Equal Protection Clause. Over the next sixteen years, the Warren Court reshaped American law: one person, one vote in the reapportionment cases; the right to counsel in Gideon v. Wainwright; the Miranda warnings; and major expansions of free speech and the rights of criminal defendants.

The significance of October 5, 1953 is a reminder of how much depends on who sits on the Court when a term opens. A recess appointment made to fill a seat for opening day put in the center chair the justice who would lead one of the most consequential eras in the Court’s history. And today, exactly 73 years later, is again the first Monday in October, with the Court opening a new term and hearing its first case: our lead story.

The Supreme Court opens its new term today by hearing arguments in one of the most important climate cases it has taken up: Suncor Energy v. County Commissioners of Boulder County. In 2018, the city and county of Boulder sued Suncor and ExxonMobil in Colorado state court under state law, alleging that the companies deceived the public about the role of their products in climate change and seeking compensation for local harms like wildfire, heat, and flooding costs. The Colorado Supreme Court ruled 5 to 2 last year that the case could go forward. The oil companies’ argument is preemption. They say that climate change is caused by greenhouse gas emissions from every country on earth, that the Clean Air Act and the federal structure govern interstate and international air pollution, and that a single state can’t use its own tort law to effectively regulate a global problem or impose liability for emissions far outside its borders. Boulder’s answer is that it isn’t regulating emissions at all. It’s suing over deception, a classic state-law claim, and nothing in federal law clearly displaces that. The Court also asked the parties to address threshold questions about whether it has jurisdiction to hear the case at this stage. Two things raise the stakes. First, nearly 60 similar suits by states and cities around the country turn on how the Court resolves this question. A broad ruling for the companies could end most of them; a ruling for Boulder would send them toward trial. Second, as we covered last week, Justice Alito has recused himself, so only eight justices are sitting. If they split 4 to 4, the Colorado ruling would stand, but without setting any national precedent, and the same fight would return in another case. A decision is expected by mid-2027.

US Supreme Court weighs bid by oil companies to avoid climate lawsuit | Reuters · Earthjustice · Harvard Environmental & Energy Law Program

The administration is asking the First Circuit Court of Appeals in Boston to revive its effort to terminate roughly $2.2 billion in federal research grants to Harvard University. The background: in April 2025, the administration sent Harvard a list of demands, including ending its diversity programs and bringing in outside auditors to monitor academic departments for “viewpoint diversity.” Harvard’s president, Alan Garber, refused, saying no government should dictate what private universities can teach. Within hours, the administration froze Harvard’s grants, citing the university’s alleged failure to address harassment of Jewish students, and it eventually terminated them. A federal district judge ruled for Harvard in an 84-page decision, finding that the administration used antisemitism as “a smokescreen for a targeted, ideologically motivated assault” on the country’s leading universities. That’s a First Amendment retaliation finding: the government can’t cut off a benefit to punish someone for protected speech or for refusing to give up academic independence. On appeal, the Justice Department makes two kinds of arguments. On the merits, it says the government isn’t required to keep funding universities that show “deliberate indifference” to antisemitism. It also argues the case was in the wrong court entirely. Under the Tucker Act, contract claims seeking money from the federal government belong in the Court of Federal Claims, not a district court, and the administration says Harvard’s suit is really a dispute over grant contracts. That jurisdictional argument has worked for the government in other grant-termination cases, so it’s a serious threat to Harvard’s win. The significance goes well beyond one university. This appeal will help decide whether the federal government can use research funding as leverage to force changes in how universities teach, hire, and govern themselves, and whether courts will review that leverage as a speech problem or treat it as an ordinary contract dispute.

US appeals court weighs Trump’s block on Harvard research funding | Reuters · US News · Harvard Magazine

And finally, Poland’s competition regulator, known by its Polish initials as UOKiK, has charged Google with abusing its dominant position in negotiations with Polish news publishers over payment for their content. The backstory is a 2024 amendment to Polish copyright law implementing an EU directive that gives press publishers a right to be paid when online platforms use their content. In Google’s case, that covers articles and snippets shown in Search, Google News, and Discover. The regulator’s theory isn’t that Google refused to pay. It’s that Google controlled the information needed to negotiate. According to UOKiK, Google didn’t give publishers the data they needed to evaluate its offers, such as how much traffic and value their content generated, so publishers had no real way to judge whether the payment was fair. The regulator said that made the negotiations “illusory,” with Google effectively imposing terms. Its president put it bluntly: big tech companies “cannot place themselves above the law.” This is an interesting antitrust theory, because the abuse alleged is an information imbalance rather than a classic refusal to deal or exclusionary practice. A legal right to payment means little if one side can’t see what the content is worth. A few points of perspective: these are charges, not a final decision, and Google will have a chance to respond. If the regulator finds a violation, the maximum fine is 10% of turnover. And this is separate from the European Commission’s investigation, opened last December, into whether Google used publishers’ content in its AI services without adequate payment or the ability to opt out. Together with the U.S. ad-tech remedy we covered last month, it’s another sign that regulators on both sides of the Atlantic are focused on the economic relationship between Google and the publishers whose content it depends on.

Polish regulator suspects Google abused dominant position in publisher payment talks | Reuters · Brandsit · Global Banking & Finance Review

This Day in Legal History: Thurgood Marshall Joins the Court

On October 2, 1967, Chief Justice Earl Warren swore in Thurgood Marshall as an associate justice of the Supreme Court, the first Black justice in the Court’s history. President Lyndon Johnson had nominated him to replace Justice Tom Clark, and the Senate confirmed him that August by a vote of 69 to 11.

Marshall arrived at the Court having already changed it from the other side of the bench. As chief counsel of the NAACP Legal Defense and Educational Fund, he argued 32 cases before the Supreme Court and won 29 of them, including Brown v. Board of Education. He had also served as a federal appeals judge and as Solicitor General. Few justices have come to the Court with a record of argument that shaped as much of the law they were about to apply.

In nearly 24 years on the Court, Marshall was a consistent voice for equal protection and for the rights of criminal defendants. He was also among the Court’s most persistent opponents of capital punishment. In Furman v. Georgia in 1972 he concluded the death penalty was unconstitutional in all circumstances, and after the Court allowed executions to resume he dissented from death sentences for the rest of his tenure. That position is worth recalling today, because our first story is about what happens when the state’s attempt to carry out a death sentence fails.

We have an update on the story we covered yesterday. Tennessee Governor Bill Lee has called the failed execution of Christa Pike “a tragedy,” saying no one wanted it to happen and that it was “deeply disturbing” that it happened in his state. As a reminder, Pike was convicted of the 1995 murder of a fellow Job Corps student, committed when Pike was 18. On Wednesday night, after the Supreme Court lifted a last-minute stay, officials administered two doses of pentobarbital. She lost consciousness but kept a steady heartbeat, and she was taken to a hospital for emergency treatment. Lee has suspended the remaining execution scheduled for this year and ordered a third-party review. The new detail is that, according to reporting from JURIST and the Washington Post, this is the second botched execution in Tennessee this year. That matters legally. Under Baze v. Rees and Glossip v. Gross, an inmate challenging an execution method generally has to show a substantial risk of severe pain and identify a feasible alternative. Courts have often treated a single mishap as an isolated accident that doesn’t prove the method is unconstitutional. Two failures in one year is harder to call isolated, and it gives lawyers for other death-row prisoners evidence that the state’s protocol itself is unreliable.

Tennessee has been here before: in 2022, Governor Lee paused executions after the state was found not to have properly tested its lethal-injection drugs. For Pike herself, the hardest question remains whether the state may try again. The 1947 case of Louisiana ex rel. Francis v. Resweber allowed a second attempt after a failed electrocution, but that was a narrow, divided decision from a very different era. Pike’s lawyers will argue a second attempt would be cruel and unusual, and they’ll keep pressing the claim that evidence of her childhood abuse was never properly considered at sentencing. Expect the third-party review, and whatever it finds about the drugs and procedures, to be central evidence in that litigation.

Tennessee governor calls failed Christa Pike execution a ‘tragedy’ | Reuters · JURIST · Washington Post

A federal judge in Miami has held that the 1996 federal law making it a crime for noncitizens to vote in federal elections is unconstitutional. Two weeks ago we covered the novel defense strategy behind this: several noncitizens charged under the law argued that Congress never had the power to pass it. At that point one Miami judge had rejected the argument. Now U.S. District Judge David Leibowitz has accepted it, dismissing the case against Chelsea Cox, a Jamaican national living in Florida charged with voting illegally in the 2020 election. His reasoning rests on the structure of the Constitution. Article I and the Seventeenth Amendment tie voter eligibility in federal elections to the qualifications each state sets for its own legislature. The Elections Clause gives Congress power over the “times, places and manner” of federal elections, but the Supreme Court said in Arizona v. Inter Tribal Council in 2013 that this doesn’t include power to set who is qualified to vote. On that view, deciding who may vote belongs to the states, and Congress can’t criminalize voting by people the Constitution leaves the states to regulate. The Justice Department’s answer, as we discussed, is that the law is really an exercise of Congress’s broad immigration power, not an election rule. Judge Leibowitz didn’t accept that framing. A few points of perspective. This appears to be the first time any federal court has found the 1996 law unconstitutional. The ruling binds only this case, and other judges, including one in the same courthouse, have gone the other way. Noncitizen voting remains illegal under the laws of essentially every state, so this doesn’t make it lawful; it questions whether the federal government can prosecute it. But the Justice Department has charged at least 60 people under this statute since January 2025, and about 45 cases are pending. Every defense lawyer in those cases now has a written federal opinion to cite. An appeal to the Eleventh Circuit seems very likely, and with judges already split, this could reach the Supreme Court.

US judge finds law criminalizing noncitizen voting unconstitutional | Reuters · Al Jazeera · GV Wire

And finally, the penalty phase of New Mexico’s case against Meta, which we covered on Monday after a Santa Fe jury found the company misled the state’s residents about Facebook’s data practices in the Cambridge Analytica case. The jury found more than 43 million violations of New Mexico’s Unfair Practices Act. The statute allows up to $5,000 per violation, which puts the theoretical maximum at about $219 billion. At a hearing Thursday, New Mexico asked Judge Francis Mathew to order Meta to pay between $35 billion and $40 billion. The state’s lawyer noted that an award that size would be the largest verdict in American legal history. Meta called the request “astronomical” and asked the judge to cap the penalty at $3.45 billion. So the judge is choosing somewhere in a range that spans more than a factor of ten. This is the legal problem I flagged on Monday: per-violation penalties were written with individual transactions in mind, and they produce enormous numbers when multiplied across a whole state’s population of users. The judge has discretion under the statute, and he’ll likely weigh things like how deliberate the deception was, how long it lasted, Meta’s ability to pay, and what amount would actually deter. There’s also a constitutional ceiling in the background. The Eighth Amendment’s Excessive Fines Clause applies to civil penalties paid to the government, and the Supreme Court held in Timbs v. Indiana in 2019 that it binds the states. A penalty grossly disproportionate to the offense can be struck down, and Meta will almost certainly make that argument on appeal if the number is large. Judge Mathew says he expects to rule later this month. Whatever he decides will be watched closely by every state attorney general with a consumer-protection case against a tech platform, because it’ll show how far these statutes can be pushed.

New Mexico wants Meta to pay up to $40 billion in penalties after data privacy trial | Reuters · Bloomberg Law · Law360

This Day in Legal History: The Nuremberg Verdicts

On October 1, 1946, the International Military Tribunal at Nuremberg delivered its sentences on the surviving leaders of Nazi Germany. Of the 22 defendants, 19 were convicted and three were acquitted. Twelve were sentenced to death, three to life imprisonment, and four to prison terms of 10 to 20 years. Hermann Göring killed himself the night before his scheduled execution. Martin Bormann, tried in absentia, was never found to face his sentence. The other death sentences were carried out on October 16.

The legal achievement of Nuremberg was the decision to hold a trial at all. The Allies could have simply executed the Nazi leadership; some officials argued for exactly that. Instead, at the urging of figures like U.S. Justice Robert Jackson, who left the Supreme Court to serve as chief prosecutor, they built a court, allowed defense counsel, admitted evidence, and let the outcome turn on proof. The acquittals mattered as much as the convictions: they showed the tribunal was deciding cases, not ratifying a foregone conclusion. Nuremberg also established principles that still anchor international law, most importantly that “following orders” is not a defense to crimes against humanity, and that individuals, not just states, can be held criminally responsible for aggression and atrocities.

The significance of October 1, 1946 is the idea in today’s opening quote, from Jackson’s opening statement: that even when dealing with the worst crimes imaginable, the state submits to legal process rather than acting on raw power. That idea, that the government must work through courts and procedure, especially when it is detaining, removing, or putting people to death, runs through every story we cover today.

The Supreme Court has agreed to decide whether the administration can detain millions of immigrants for the entire length of their deportation proceedings without a bond hearing. In July 2025, the Department of Homeland Security issued guidance saying that immigrants who entered the country illegally are subject to mandatory detention with no bond hearing while their cases proceed, a process that can take months or years. The legal fight is about which section of the immigration statute applies. The law has one provision that requires detention for “applicants for admission,” historically understood to mean people arriving at the border. A separate provision governs people already living in the country and generally lets them ask an immigration judge for release on bond. The administration’s new position is that anyone who entered without inspection remains an “applicant for admission” no matter how long they’ve lived here, which would move them into the mandatory-detention category. The case the Court took involves Ricardo Aparecido Barbosa da Cunha, a Brazilian national who authorities say entered illegally two decades ago and applied for asylum in 2016. The Second Circuit ruled against the government. And according to the reporting, appeals judges in nine of the eleven federal circuits to consider the question have concluded that detention without a bond hearing violates federal law. That makes the administration’s position an outlier among the lower courts, which is part of why the Supreme Court stepped in. The stakes are very large. A bond hearing doesn’t guarantee release; it just means a judge decides whether a person is a flight risk or a danger. If the Court sides with the administration, millions of long-term residents could be held without that individualized review for the length of their cases. The Court will hear it in the term that opens Monday.

US Supreme Court to hear clash over Trump’s immigration detention policy | Reuters · Yahoo News · National Immigration Forum

The U.S. Court of International Trade heard arguments in a challenge to the President’s “forced labor” tariffs, and opponents say the case is a test of whether he can rebuild the tariff power the Supreme Court took away in February. Imposed in late July, the tariffs range from 10% to 12.5% on goods from 60 trading partners, including the European Union and China, and they cover more than 99% of goods imported into the United States. The stated justification is that those countries have failed to stop imports made with forced labor. Four small businesses and 25 Democratic-led states sued, and their three cases were consolidated before a three-judge panel with appointees of Presidents Trump, Obama, and Biden. Their argument is that the trade statute the administration relied on requires country-specific findings of unfair trade practices, reached through a real investigation. In their telling, “forced labor” is a pretext: a label stretched over nearly all imports from 60 countries to recreate the broad, global tariff authority the Supreme Court rejected when it struck down the emergency-powers tariffs. The legal question is how closely courts will examine the factual basis for a statutory trade remedy. If the statute requires findings and the findings are thin or generic, the tariffs are vulnerable. If the court defers to the executive’s judgment on trade practices, they likely survive. The panel isn’t expected to rule from the bench. This connects directly to my column from last week on building an unwinding mechanism into tariffs from the start. If these tariffs fall too, importers will be back in the refund line, and consumers who paid higher prices will again have no clear way to get that money back.

US trade court to weigh challenge to Trump’s forced labor tariffs | Reuters · CNBC · Foreign Policy

And finally, a story that took a turn no one expected. Tennessee tried to execute Christa Pike on Wednesday night, and the execution failed. Pike was the only woman on Tennessee’s death row, convicted of the 1995 torture and murder of 19-year-old Colleen Slemmer, a fellow student at a Knoxville Job Corps program. Pike was 18 at the time of the crime. The legal path to Wednesday night was itself dramatic. On Wednesday morning, a divided panel of the Sixth Circuit granted a short stay to consider her claim that evidence of severe childhood sexual abuse was never adequately considered at sentencing. Tennessee went to the Supreme Court, which vacated the stay Wednesday evening, with Justices Sotomayor, Kagan, and Jackson dissenting. Officials then administered two doses of the lethal drug, but Pike’s heart was still beating, and she was taken to a hospital. Governor Bill Lee has ordered a comprehensive third-party review and said the remaining execution scheduled for this year will not go forward. Now the legal questions are unusual and difficult. The closest Supreme Court precedent is from 1947, Louisiana ex rel. Francis v. Resweber, when the state’s electric chair failed to kill Willie Francis and the Court held, 5 to 4, that Louisiana could try again without violating the Eighth Amendment. Modern lethal-injection cases like Baze v. Rees and Bucklew v. Precythe focus on whether a method creates a substantial risk of severe pain, but they don’t squarely address what happens after an execution has actually been attempted and failed. Pike’s lawyers will almost certainly argue that a second attempt would be cruel and unusual. There are also the issues that were already in the case: her age at the time of the crime, just past the line the Supreme Court drew in Roper v. Simmons barring execution for crimes committed under 18, and the abuse evidence the Sixth Circuit wanted time to examine. Whatever one’s view of the death penalty, a failed execution is exactly the kind of event the Eighth Amendment was written to make courts confront, and this case will now test what the Constitution requires when the state’s ultimate punishment doesn’t work.

Tennessee due to execute woman for first time in two centuries | Reuters · NPR · CBS News

This Day in Legal History: The Integration of Ole Miss

On September 30, 1962, James Meredith, a Black Air Force veteran, was escorted onto the campus of the University of Mississippi by federal marshals to enroll under a federal court order. The university had rejected him for no reason other than his race. The Fifth Circuit ordered him admitted, and Justice Hugo Black refused to let that order be delayed. Mississippi’s governor, Ross Barnett, personally blocked Meredith’s registration anyway. The federal appeals court held him in contempt.

After nightfall on the 30th, a crowd of students, locals, and segregationists attacked the roughly 120 marshals guarding Meredith. That night President Kennedy went on national television to explain why the federal government would enforce the order, the address today’s opening quote comes from. The riot killed two people, including a French journalist, and injured hundreds. It took thousands of federal troops to end it. The next morning, Meredith registered.

Listeners may remember we covered Little Rock earlier this month. Ole Miss is the same principle five years later: a federal court order is binding, even on a state determined to resist it, and the federal government will enforce it. The Constitution’s guarantees mean what courts say they mean, and a governor doesn’t get to veto them. That idea of judicial orders as binding law, not suggestions, has come up again and again on this show this month.

The Supreme Court has let the administration resume deporting migrants to so-called third countries: nations that aren’t named in their removal orders and that many of them have no connection to. In a brief order, with the three liberal justices objecting, the Court paused a February ruling by a federal district judge. That ruling required immigration authorities to give people more due process before sending them to a third country, and a federal appeals court had largely upheld it earlier this month. The administration says it has used the policy to deport tens of thousands of people to Mexico and to countries in Africa and Latin America, often after their home countries refused to take them back. The legal issue is notice and an opportunity to be heard. The challengers argue that if the government plans to send someone to a country where they may face persecution or torture, due process and the Convention Against Torture require that the person be told where they’re going and get a real chance to object first. The government’s position is that it has broad authority over removals and that lower courts overstepped. The Court also agreed to hear the case in full in December, and it asked the parties to brief several questions: whether the lower courts had jurisdiction at all, whether they could order nationwide relief, and whether the policy violates due process or the torture convention. Those first two questions matter well beyond immigration, because they go to the reach of federal judges’ power over executive policy. For now, the deportations resume while those questions are decided. As with the SAVE voter-verification order last week, the emergency docket has put the risk of error on the people affected while the case is pending.

Supreme Court lets Trump resume third-country deportations | Reuters · NPR · CBS News

Former special counsel Jack Smith testified before the Senate Judiciary Committee at a hearing its Republican chairman, Chuck Grassley, titled an oversight hearing into Smith’s “abuse of authority.” Smith brought the two federal criminal cases against Donald Trump: one over his efforts to overturn the 2020 election, and one over his handling of classified documents after his first term. Both cases were dismissed after Trump won the 2024 election. Grassley accused Smith of “rampant abuse of authority and political targeting” funded by taxpayers. Smith defended the investigations as based on evidence and Justice Department policy, not politics. He told senators he would not be silenced by “continued threats of prosecution,” and he condemned the firing and vilification of the agents and prosecutors who worked for him. Live coverage also reported him saying his investigation had shown, beyond a reasonable doubt, that Trump engaged in serious crimes. Here’s the legal significance. A special counsel operates under Justice Department regulations designed to give some independence in politically sensitive investigations. When a former prosecutor faces open talk of prosecution for his own charging decisions, and his former staff are fired, the concern is less about Smith himself than about future prosecutors. Every prosecutor considering a case against a powerful person will now know that the charging decision itself can bring personal and professional consequences. Congress does have legitimate oversight power over the Justice Department. But oversight turning into retaliation is exactly what Justice Department independence norms are supposed to prevent. The hearing is another sign of how politicized federal prosecution has become.

Former Trump prosecutor Jack Smith faces Senate hearing over abuse of authority accusations | Reuters · NBC News · CNN

And finally, in my column for Bloomberg Tax this week, I take on one of the most common arguments against California’s proposed wealth tax. Proposition 40 would impose a one-time 5% levy on the net worth of Californians who were billionaires at the start of this year. Opponents warn that a tax on the ultra-rich will eventually trickle down to ordinary households. My response is that ordinary households already pay a recurring tax on their largest store of wealth. It’s called the property tax. Think about how it works: the government values an asset you own, applies a rate to that value, and bills you every year simply for continuing to own it. That’s a tax on a stock of wealth, not on income. I ran some back-of-the-envelope numbers using Tax Foundation and Federal Reserve data. The national effective property tax rate on owner-occupied homes was about 0.91% in 2022. The median primary residence was worth about $323,000, and median housing equity was about $200,000. That works out to a property tax bill of roughly $2,900 a year, or close to 1.5% of the typical homeowner’s actual equity, every year. And unlike Proposition 40, the property tax doesn’t subtract your mortgage. It’s charged on the full assessed value, so with a big mortgage, a nominal 1% rate can equal 5% or more of your real equity. The billionaire’s wealth tax would be calculated on assets minus debts, which is more lenient on that point. My conclusion is that many homeowners would gladly trade their recurring 1.5% for a one-time 5%. So the real question isn’t whether taxing wealth is acceptable in principle. It’s why our appetite for it changes so much depending on what kind of wealth it is and who owns it. I’m not saying Proposition 40 is well designed. Valuing private company shares is much harder than valuing a house, and billionaires can move in a way houses can’t. If the objection is really to the size and one-time structure, the better debate is whether a recurring, lower-rate tax, say 1% a year above a very high threshold, would be workable. That would look much more like the property tax system homeowners already live with, and over ten years it would raise roughly double the revenue.

California Wealth Tax Trickle-Down Concerns Are Unconvincing | Bloomberg Tax

This Day in Legal History: Congress Creates the U.S. Army

On September 29, 1789—the very last day of the First Congress’s inaugural session—the House and Senate passed the act that formally established the United States Army under the new Constitution. President Washington had pressed hard for it, reminding Congress that the militia provisions inherited from the Articles of Confederation were inadequate and that the nation needed a standing military establishment placed on a proper constitutional footing.

The legal significance lies less in the creation of a fighting force than in how that force was subordinated. The Constitution had already divided military power deliberately: Congress alone can raise and fund armies and declare war, while the President serves as Commander-in-Chief. The 1789 Act operationalized that division, embedding the bedrock American principle of civilian control of the military—that the armed forces answer to elected civilian authority and operate within, not above, the rule of law. The Framers were deeply wary of standing armies and concentrated power; their answer was not to forbid a military, but to fracture control over it between the branches so no single actor could wield it unchecked.

The significance of September 29, 1789 is that principle—power channeled through congressional authorization and constitutional limits rather than executive will alone. And that is precisely the thread running through our lead story today: a dispute over whether the executive branch can wield federal money as a lever to force its policy preferences on the states, or whether it may only do what Congress has actually authorized. The names change over 237 years; the fundamental question—who gets to exercise power, and with whose permission—does not.

A federal judge has blocked the administration from tying federal counterterrorism grants to whether states adopt its preferred election-administration changes—the latest in a long string of election-related defeats we’ve tracked this month. Here’s what the administration did: the Department of Homeland Security announced it would withhold 20% of funding under grant programs meant to help state and local governments prepare for terrorist attacks and other threats, unless those governments adopted election measures the administration favored. U.S. District Judge Amir Ali, in a 24-page opinion, sided with counties and cities in Texas, Ohio, and Tennessee that sued, holding that federal agencies simply lacked authority from Congress to attach those conditions to the grants. Let me put this in its constitutional frame, because it’s a classic. The federal government routinely attaches strings to the money it gives states—that’s the spending power, and it’s legitimate within limits. But the Supreme Court set boundaries in cases like South Dakota v. Dole and, more forcefully, in the NFIB v. Sebelius healthcare decision: conditions must be clearly stated, must relate to the purpose of the funding, and can’t be so coercive that they amount to a “gun to the head.” Conditioning anti-terrorism preparedness money on unrelated election changes runs into two of those problems at once—the germaneness requirement, since election administration has nothing to do with counterterrorism, and the basic separation-of-powers point that an agency can’t invent conditions Congress never authorized. That last point is what Judge Ali hung his ruling on, and it’s the same principle that decided the homelessness and mail-ballot cases: agencies have only the power Congress gives them. The significance is that this is now a consistent, rulings-deep pattern—court after court telling the executive it cannot use funding leverage, or agency fiat, to seize control over how states run their elections. On the anniversary of a law about channeling power through congressional authorization, it’s a fitting rebuke.

Judge blocks Trump from tying anti-terrorism grants to election changes | Reuters · PBS News · US News

In a notable reversal, Justice Samuel Alito has recused himself from a major climate-change case the Supreme Court is set to hear next week—and the move opens up real questions about both judicial ethics and the case’s outcome. The case involves Boulder, Colorado, and its county, which sued Exxon Mobil and Suncor Energy seeking damages for the local harms of climate change; it’s set for argument October 5. Environmental groups had pushed for Alito to step aside because he owns stock in oil-and-gas companies—ConocoPhillips and Phillips 66, among others—that, while not parties to this particular suit, plainly have interests riding on how the Court treats climate liability. Here’s why the reversal is striking. Back in May, a Court spokeswoman said Alito would not recuse, on the theory that he had no financial interest in any actual party to the case. Now he’s stepping aside anyway. This goes to the heart of the judicial-ethics debate that has dogged the Court: federal law, 28 U.S.C. § 455, requires a judge to disqualify when their impartiality might reasonably be questioned or when they have a financial interest in the subject matter in controversy—and Supreme Court justices, unlike lower-court judges, have historically decided their own recusal questions with no review and little explanation. Alito’s flip suggests the pressure and scrutiny over the Court’s ethics, which produced its first-ever formal code of conduct in 2023, is having an effect. And there’s a concrete consequence: with only eight justices participating, the case could end in a 4-4 tie—which would leave the lower-court ruling in place without setting any national precedent. The significance is that it’s a small but real sign of the Court responding to ethics scrutiny, and a reminder that in the climate-liability wars, procedural moves like a single recusal can matter as much as the merits.

US Supreme Court’s Alito won’t participate in climate case involving oil companies | Reuters · CNBC · NBC News

And finally, Florida has asked a court to bar OpenAI from developing new AI models without outside oversight—an extraordinary requested remedy that, if granted, would be one of the most aggressive judicial interventions into AI development we’ve seen. This is part of a lawsuit Florida’s attorney general, James Uthmeier, filed in June, accusing OpenAI of misrepresenting ChatGPT’s safety and of harming children—the complaint alleges the platform provided information to school shooters, offered guidance on self-harm, and was designed to addict young users. In Monday’s filing, Florida escalated the requested relief dramatically: it asked the court to prohibit OpenAI from developing new models without external oversight, to order the company to keep minors off ChatGPT entirely, and to bar it from giving the chatbot “human attributes.” Let me unpack why that’s such a big ask. Courts issue injunctions all the time, but a prohibitory injunction against building new products—essentially putting a court-supervised brake on a technology company’s core R&D—is a remarkable expansion of the injunctive remedies we’ve discussed in the teen-harms cases. It raises hard questions about how far a court’s equitable power can reach into ongoing innovation, and whether such sweeping prior restraint on development is even administrable. Notably, OpenAI says it has already paused training its most capable models and won’t resume until additional safeguards are in place—which both undercuts the urgency Florida claims and signals the company sees the writing on the wall. Florida is the first state AG to sue OpenAI over harms to young users, though the company faces individual and family suits too, including ones tied to self-harm. And this connects to a huge thread we’ve followed all month—the Harvey guardrails acquisition, the ChatGPT-fabricated testimony, California’s Adam’s Law. The significance is that the legal system is now being asked not just to punish AI harms after the fact, but to preemptively supervise how these models get built—a genuine frontier for the law, and one courts will approach with real caution.

Florida asks court to bar OpenAI from developing new models as part of child harm lawsuit | Reuters · US News · Lawyer Monthly

This Day in Legal History: The Black Sox Indictment

On September 28, 1920, a Chicago grand jury indicted eight members of the Chicago White Sox for conspiring to throw the 1919 World Series in exchange for money from a gambling syndicate. That morning, pitcher Eddie Cicotte and “Shoeless” Joe Jackson had reportedly confessed. The scandal—forever after known as the “Black Sox”—remains one of the most famous intersections of crime, sport, and law in American history, and its legal legacy is more instructive than the folklore suggests.

Here’s the twist that every law student should sit with. The eight players went to trial on criminal conspiracy charges, and in 1921, a jury acquitted all of them. But the very next day, baseball’s newly installed commissioner, federal judge Kenesaw Mountain Landis, banned all eight from organized baseball for life—famously declaring that regardless of any jury verdict, no player who fixes games or consorts with gamblers would ever play professional baseball again. So you have a stark divergence: the criminal-justice system, applying its high beyond-a-reasonable-doubt standard, found them not guilty, while a private governing body, applying its own rules and lower burden of proof, imposed the harshest sanction it could.

The significance of September 28, 1920 is that lesson in parallel systems of accountability. Being cleared in a court of law is not the same as being cleared everywhere—private institutions, licensing bodies, and employers routinely impose their own consequences under their own standards, and they’re generally entitled to. It’s a theme that runs straight through today’s episode, where we’ll see companies held to account not by criminal courts but by state consumer-protection laws and civil juries—the many rooms, beyond the criminal courthouse, in which accountability actually happens.

A jury in Santa Fe has found that Meta misled New Mexico residents about how it protected their data—and the potential penalties are genuinely staggering. This case traces back to the Cambridge Analytica scandal: the 2018 revelation that the political consulting firm, tied to the 2016 Trump campaign, harvested personal data from as many as 87 million Facebook users through a third-party app without their consent. New Mexico’s attorney general sued in 2021, and after a two-week trial, the jury found that 26 of 29 statements Meta made were misleading, adding up to tens of millions of violations of the state’s Unfair Practices Act—affecting essentially the entire population of the state, more than two million people. Here’s where it gets eye-popping. The judge, not the jury, will now set penalties, and New Mexico is seeking the statutory maximum of $5,000 per violation. Do the math on tens of millions of violations and you get a theoretical exposure reaching into the billions—some coverage has floated figures over $200 billion, though the actual number the judge lands on will almost certainly be far lower, because courts have discretion and because due-process principles cap grossly excessive aggregate statutory penalties. That’s the real legal drama here: the collision between per-violation statutory damages, designed for individual harms, and mass digital conduct affecting millions, where mechanical multiplication produces astronomical, arguably unconstitutional numbers. The significance is twofold. First, it’s a powerful demonstration of state consumer-protection law as a weapon against Big Tech—New Mexico didn’t need a federal privacy statute, it used its own deceptive-practices act. And second, it feeds directly into the trend we keep tracking: after years of Cambridge Analytica settlements, a jury has now actually found Meta liable to a state’s entire population for deceiving them about privacy. Meta says it disagrees and will keep fighting.

Meta misled consumers in case over Cambridge Analytica scandal, New Mexico jury says | Reuters · PBS News · Engadget

TikTok has settled with Alabama for at least $100 million—and up to $300 million if certain conditions are met—just days before what would have been the first state trial in the country over the platform’s alleged harms to teenagers. This is a significant development in a fight we’ve followed closely. At least 27 other states and D.C. have sued TikTok on similar theories: that it deliberately designed its platform to be addictive to children and misled the public about how safe it is. Alabama’s case was set to go before a Montgomery jury this Monday, in a trial expected to run two to three weeks and to pry open TikTok’s internal workings—which is exactly the kind of exposure that tends to concentrate a company’s mind on settling. And this settlement is notable not just for the money but for the injunctive relief, because the platform changes are substantial: overnight access restrictions for teen users between midnight and 6 a.m., limits on messaging and notifications overnight and during school hours, stronger age-verification, a ban on cosmetic filters for teens, a default non-personalized feed, beefed-up parental controls, and measures to keep adults from discovering teen accounts. From a legal standpoint, that injunctive component is arguably the more important half—money is a one-time cost a company TikTok’s size can absorb, but forced design changes go to the heart of the business model the lawsuits target. It also connects to the whole ecosystem we’ve covered: the Texas ruling that TikTok misled users on child safety, the Meta addiction cases, California’s Adam’s Law. The significance is that TikTok, facing the prospect of airing its internal documents before a jury, chose to pay and change its product rather than risk a verdict—and it sets an expensive template that the other 27-plus states will now be measuring their own cases against.

TikTok settles with Alabama ahead of first state trial over claims of teen harms | Reuters · CNBC · US News

And finally, a major reversal in an arc we’ve tracked all month: the Supreme Court has restored the administration’s mass voter-verification system, clearing states to use it ahead of the November midterms. In an unsigned emergency order, with the Court’s three liberal justices dissenting, the justices blocked a lower-court order that had stopped officials from using the Department of Homeland Security’s revamped SAVE system—the Systematic Alien Verification for Entitlements database—to run sweeping citizenship checks against voter rolls. This is the very system whose block we celebrated as upheld back in early September, when the D.C. Circuit kept it frozen; now the Supreme Court has lifted that freeze, at least while the litigation continues. Recall the design: at the President’s March 2025 direction, DHS linked the 40-year-old SAVE program with Social Security records to enable bulk citizenship checks identifying supposed noncitizen voters. And recall the core objections, which haven’t gone away. First, privacy—the League of Women Voters and the Electronic Privacy Information Center argue the revamp violates federal privacy laws by exposing Social Security data. Second, and most alarming, accuracy—investigations in Travis County, Texas, found that somewhere between 10% and 21% of voters flagged by SAVE were actually U.S. citizens. Sit with that: a tool being used to police voter rolls weeks before an election that may wrongly flag up to one in five as noncitizens, forcing real citizens to prove their status or risk removal. Now, an emergency-docket order like this is not a ruling on the merits—it’s a decision about who bears the risk while the case proceeds, and the 6-3 split tells you the majority was willing to let the system run for now. The significance is heavy: after a month of courts repeatedly blocking federal efforts to centralize control over voter rolls, the Supreme Court has, on the shadow docket and over dissent, let the most sweeping of those tools go live for the midterms—accuracy concerns and all.

Supreme Court restores Trump’s mass voter verification system | Reuters · NPR · CNBC

This Day in Legal History: Congress Proposes the Bill of Rights

On September 25, 1789, the First Congress agreed on twelve proposed amendments to the newly ratified Constitution and sent them to the states for ratification. Ten of those twelve would be ratified by the states in December 1791, and we know them today as the Bill of Rights. It was the fulfillment of a promise—many states had ratified the Constitution only on the understanding that a bill of rights protecting individual liberties would swiftly follow.

The amendments are the beating heart of American individual liberty. The First Amendment’s protections for speech, press, religion, and assembly; the Fourth Amendment’s shield against unreasonable searches; the Fifth and Sixth Amendments’ guarantees of due process and fair trials; the Eighth’s ban on cruel and unusual punishment. What’s remarkable is how the meaning of these spare, elegant clauses has been fought over and elaborated for more than two centuries—the words are fixed, but their application to new circumstances is a never-ending project. It’s worth remembering, too, that two of the original twelve weren’t ratified in 1789: one concerning congressional apportionment, and one on congressional pay—and that pay amendment was finally ratified in 1992, more than 200 years later, as the 27th Amendment.

The significance of September 25, 1789 is that it created the framework of enumerated rights that Americans invoke against their government every single day—and never more directly than in today’s lead story, which is, at bottom, a First Amendment fight over freedom of the press playing out at the doors of the White House. On the anniversary of the day those freedoms were proposed, we watch a court try to enforce them in real time.

We have a remarkable follow-up to a story we covered earlier this week: the White House barred reporters from CNN, MS NOW, and Politico from entering—just hours after a federal judge ordered their access restored and deemed the ban likely unconstitutional. Let me lay out the sequence, because it’s striking. Recall those three outlets sued after President Trump banned them over what he called their “negative stories.” This week, U.S. District Judge Timothy Kelly ruled for the outlets, temporarily blocking the ban for 14 days and finding it likely unconstitutional—a clear First Amendment win. And then, hours later, when reporters from those outlets tried to enter the White House grounds, Secret Service officers turned some of them away and even confiscated a Politico reporter’s credentials. The outlets rushed back to court, requesting an immediate hearing, and by the end of the day the White House began readmitting the journalists. Now, here’s the legal significance, and it’s serious, because this brushes up against one of the gravest issues in our legal system: compliance with court orders. When a court issues an injunction, parties are legally bound to obey it—full stop. The rule of law depends on the understanding that when a judge rules, even the most powerful actors, including the executive branch, comply. When there’s a gap—even a few hours—between a court order and the government’s compliance with it, that’s not a small administrative hiccup; it raises the specter of defiance of the judiciary, which strikes at the foundation of the separation of powers. Now, the charitable read is that this was a logistical lag—word hadn’t reached the officers at the gate—and the fact that access was restored by day’s end supports that. But it happened against a backdrop of rising concern about executive compliance with court orders generally. The significance is that the First Amendment merits question—can a President ban outlets for critical coverage—now sits alongside an even more fundamental one: whether court orders enforcing those rights will be promptly obeyed. On the anniversary of the Bill of Rights, that’s a sobering pairing.

Politico denied access to White House after judge’s order, news outlet says | Reuters · Washington Post · CNBC

The American Bar Association is fighting to hold onto one of its most consequential powers: its role as the federally recognized accreditor of the nation’s law schools. This week, the ABA’s legal-education council was set to meet with a federal advisory committee to defend that role, after the Department of Education recommended in August that the government withdraw recognition of the ABA as the official law-school accreditor. Let me explain why this is a big deal, because accreditation is quiet but enormously powerful. In most states, you have to graduate from an ABA-accredited law school to even sit for the bar exam—so the ABA’s accreditation standards effectively shape what every law school in America must teach, how it must operate, and who can ultimately enter the profession. That’s a tremendous amount of gatekeeping authority concentrated in a private professional organization. The administration’s objection traces to an April 2025 executive order and centers on the ABA’s diversity, equity, and inclusion requirements—specifically Standard 206, which required schools to demonstrate a commitment to diversity. And here’s the telling development: earlier this month, the ABA’s council voted 10 to 6 to repeal Standard 206, the very DEI requirement at the heart of the dispute. So the ABA is fighting on two fronts at once—defending its institutional role while simultaneously giving ground on the substantive policy that drew the administration’s fire. This fits squarely within the anti-DEI campaign we’ve tracked across this administration, from the tax-exempt-status proposal to the pressure on universities. But it’s also part of something broader—a contest over who controls the pipeline into the legal profession itself. The significance is real: if the government strips the ABA of its accreditor status, it could reshape legal education, shift accreditation power to the states or a new body, and mark an extraordinary intervention by the executive branch into the self-governance of the legal profession. For a profession that prizes its independence, that’s a genuinely existential fight.

ABA fights to retain law school governance role in meeting with Trump education department | Reuters · Inside Higher Ed · Fox Business

And finally, New York has sued the prediction market Polymarket, calling it an illegal, unlicensed gambling operation—and the case cuts right to a question our legal system is scrambling to answer: what exactly is a prediction market? Attorney General Letitia James filed suit seeking to block Polymarket from operating in New York, arguing that its event contracts—where users put money on uncertain future outcomes, from elections to sports to who-knows-what—meet New York’s legal definition of gambling. And since Polymarket isn’t licensed by the New York State Gaming Commission, the state says it’s operating illegally. James raised two pointed concerns: that Polymarket lets 18-to-20-year-olds trade sports markets even though New York requires mobile sports bettors to be 21, and that it avoids the taxes that licensed sportsbooks have to pay. Here’s the fascinating legal core, and it’s a classic federalism clash. Polymarket hit back within hours with its own suit, arguing that New York is overstepping, because Polymarket operates as a federally regulated exchange—its contracts are overseen by the Commodity Futures Trading Commission, the federal commodities regulator. So the battle is: is a bet on an election outcome “gambling,” which states have always regulated, or is it a “swap” or derivative—a financial instrument—which falls under exclusive federal CFTC jurisdiction? If it’s the latter, federal law may preempt New York’s gambling regime entirely. This is the same fight James picked with Polymarket’s rival Kalshi two months ago, and courts around the country are wrestling with it. The significance is that the outcome will help define whether the booming prediction-market industry is treated as a financial market or as a casino—a multibillion-dollar question about which sovereign gets to regulate it, and under which body of law. It’s regulatory categorization as destiny, and right now nobody’s entirely sure which box these things belong in.

New York sues Polymarket, says it ran illegal gambling operation | Reuters · Bloomberg · CNBC

This Day in Legal History: The Judiciary Act of 1789

On September 24, 1789, President George Washington signed the Judiciary Act of 1789 into law—one of the very first acts of the very first Congress, and a document that turned the skeletal promise of Article III into a functioning national court system. The Constitution had created “one supreme Court” and left everything else to Congress’s imagination. The Judiciary Act supplied the imagination: it built a three-tiered federal judiciary of district courts, circuit courts, and the Supreme Court; it fixed the number of justices at six; and it created the office of Attorney General.

The Act’s most consequential provision turned out to be a quiet one: Section 25, which gave the Supreme Court the power to review and reverse state-court decisions that ran against federal law or the Constitution. That authority—federal judicial supremacy over state courts on questions of federal law—was the seed of much of what the Supreme Court would later become, and Section 13 of the same Act was the very provision the Court would strike down in Marbury v. Madison, establishing judicial review itself. In other words, the machinery of American judicial power, and the Court’s authority to police the boundary between state and federal law, traces directly to this statute.

The significance of September 24, 1789 is that it is the architecture beneath essentially everything we cover—the district courts where cases begin, the circuit courts of appeals that review them, and the Supreme Court that sits atop it all. And it’s a strikingly apt anniversary, because our lead story is that very system in action, and under strain: a dispute ricocheting between a state supreme court, a federal district judge, a federal circuit court, and the U.S. Supreme Court—the exact interplay of state and federal judicial power that the Act of 1789 first set in motion.

The fight over Missouri’s congressional map is back at the U.S. Supreme Court—for the third time in a month—and it has become a genuinely dizzying illustration of how tangled our courts can get. Let me trace the bouncing ball, because the procedural chaos is the story. Missouri Republicans adopted a new map in 2025 that dismantled a Kansas City district held by Democrat Emanuel Cleaver, part of the Trump-backed national redistricting push. Then: on September 3, the Missouri Supreme Court unanimously ruled that under state law the new map can’t be used until voters approve it in a referendum. On September 8, Justice Kavanaugh, as circuit justice, rebuffed Missouri’s request to intervene. On September 10, the U.S. Supreme Court blocked the state from using the redrawn map, in an unsigned order with no dissent. But then a federal district judge and, this Monday, the 8th Circuit Court of Appeals went the other way, reviving the Republican map—which has now triggered this third trip to the Supreme Court. Here’s the legal knot at the center: this is a collision between state law and federal law. The Missouri Supreme Court’s ruling rests on the state constitution’s referendum requirement; the federal proceedings involve federal claims about the map. Untangling which sovereign’s law controls, and which court has the final say, is exactly the state-versus-federal judicial question the Judiciary Act of 1789 first tried to sort out—and today it’s playing out in real time. Meanwhile, the human cost is real: more than a million potential voters are caught in the confusion, and absentee voting is already underway using the old 2022 districts. The significance is twofold—it’s another test of mid-decade partisan gerrymandering, and it’s a vivid, almost overwhelming example of what happens when state and federal courts point in opposite directions weeks before an election. Whatever the Court does, doing it this late, on the emergency docket, leaves election administrators and voters in an impossible bind.

Battle over Missouri’s congressional map reaches US Supreme Court for third time | Reuters · SCOTUSblog · CNN

A federal judge has dismissed Michigan’s climate lawsuit against the oil industry—and the way she did it matters, because Michigan tried a novel legal theory that just ran aground. Most of the climate suits we’ve seen from states and cities are built on public-nuisance and consumer-protection theories—the claim that fossil-fuel companies deceived the public about climate change. Michigan tried something different and more ambitious: an antitrust theory. Attorney General Dana Nessel accused BP, Chevron, Exxon, Shell, and the American Petroleum Institute of conspiring, over decades, to suppress competition from electric vehicles and renewable energy in order to preserve fossil fuels’ dominance. The idea was to reframe climate harm as an antitrust injury—collusion to kill off cleaner competitors. U.S. District Judge Jane Beckering in Grand Rapids rejected it, and her reasoning is a classic antitrust-doctrine problem: proximate cause and antitrust standing. She found that antitrust law simply doesn’t protect against most of the injuries Michigan claimed, and that even for the one cognizable category—energy overcharges—”the distance is too great between the alleged conspiracy and Michigan’s and its residents’ overcharges” to say the conspiracy actually caused them. That’s the antitrust-standing doctrine from cases in the lineage of Associated General Contractors: to sue, your injury has to be the kind antitrust law was meant to prevent, and it can’t be too remote or speculative a link down a long causal chain. The significance is that this marks a setback for a creative frontier in climate litigation. The public-nuisance suits grind on in various states, but Michigan’s attempt to weaponize antitrust law against Big Oil for slow-walking the energy transition has, at least here, been deemed too attenuated a theory to proceed. It’s a reminder that even a compelling narrative of corporate misconduct has to fit within the specific, technical boundaries of the legal theory you choose.

US judge dismisses Michigan climate lawsuit against oil companies | Reuters · Inside Climate News· US News

And finally, the Musk-versus-OpenAI antitrust brawl has taken a delicious turn: OpenAI is trying to get the case thrown out by using Elon Musk’s own SEC filings against him. Recall the posture we covered—Musk’s xAI and X Corp sued Apple and OpenAI, claiming Apple’s exclusive integration of ChatGPT into the iPhone illegally shut out rivals like Grok. Then, on September 14, Musk’s companies quietly dropped Apple from the suit, leaving OpenAI as the lone remaining defendant. Now OpenAI has asked Judge Mark Pittman in Fort Worth for summary judgment—a ruling in its favor on the existing record, before the January trial. And its argument is beautifully simple. OpenAI points to the IPO registration statement that Musk’s SpaceX filed with the SEC, and says it is “replete with disclosures diametrically opposed” to xAI’s claims of competitive harm—that the rosy, optimistic picture a company is legally required to paint for investors “bears no resemblance to the doomsaying in this litigation.” Here’s why this is legally clever, and it goes to the heart of securities law. When you file with the SEC, you are under a legal obligation to be truthful and not to mislead investors—so a company’s SEC disclosures are treated as serious, considered admissions. If SpaceX and xAI told investors the AI market is competitive and full of opportunity, they can’t easily turn around and tell a court the same market is being unlawfully monopolized to their ruin. It’s the litigation version of getting caught saying two contradictory things to two different audiences—and courts do not look kindly on it. The significance is a sharp lesson that echoes my own tax-and-regulation beat: your legally-required disclosures in one forum can come back to bind you in another. You cannot tell Wall Street one story and a federal judge the opposite. Whether it’s enough to end the case before trial is up to Judge Pittman, but OpenAI has landed a genuinely elegant punch.

OpenAI says SEC disclosures undermine xAI’s antitrust lawsuit | Reuters · PYMNTS · Unite.AI

This Day in Legal History: The Capture of John André

On September 23, 1780, three American militiamen stopped a man riding toward British lines near Tarrytown, New York. Hidden in his boot were documents and maps of the fortifications at West Point. The man was Major John André, adjutant general of the British Army—and the papers exposed one of the most infamous plots in American history: General Benedict Arnold’s secret agreement to surrender West Point to the British in exchange for money and a British command. Arnold escaped to the enemy; his name became a synonym for traitor. André was not so lucky.

The legal aftermath is what makes this a landmark. André was tried not in a civilian court but before a board of fourteen American general officers—a military tribunal—which found him to be a spy and condemned him to death. He was hanged on October 2, 1780. And here’s the part worth pausing on: André conducted himself with such dignity, and the proceedings raised such difficult questions, that his case became an enduring reference point in the law of war and the treatment of spies. Even George Washington, who approved the execution, wrestled with it. The core issues—what process is owed to someone the government has seized on national-security grounds, when a military tribunal may substitute for an ordinary court, and how a nation balances self-preservation against the procedural fairness it claims to stand for—are not relics. They run from André, through the Civil War case of Ex parte Milligan, through the World War II saboteur case of Ex parte Quirin, all the way to the post-9/11 Guantanamo litigation.

The significance of September 23, 1780 is that it poses, at the very founding of the country, the permanent question of how a nation treats those it deems dangerous to its security—and how much process it owes them even so. That question is not historical trivia today. It is, almost exactly, the question in our lead story.

The Trump administration has asked the Supreme Court to allow the detention of a pro-Palestinian scholar—a case that squarely tests whether federal immigration law can override a noncitizen’s due-process and free-speech claims. The scholar is Badar Khan Suri, an Indian citizen and postdoctoral fellow at Georgetown who was arrested by immigration authorities in March 2025, part of the administration’s broader campaign to detain and deport foreign nationals who engaged in pro-Palestinian activism on U.S. campuses. A lower-court judge ordered Suri released after he argued his detention violated his First Amendment free-speech rights and his Fifth Amendment right to due process. Now the Justice Department is asking the Supreme Court to consider whether federal immigration law actually stripped that judge of the authority to order him released in the first place. And that framing is the whole ballgame. The government’s argument is jurisdictional: it contends that provisions of the immigration laws channel these disputes away from the federal district courts and their habeas power, so the judge had no business ordering release. The defense is constitutional: that no statute can wall off the courts from hearing claims that the government is imprisoning someone to punish protected speech. This is the same fight playing out in the parallel case of Mahmoud Khalil, the Columbia activist, where a different appeals court came out the other way—setting up exactly the kind of circuit split that draws the Supreme Court in. The significance could hardly be higher. This asks whether the government can detain noncitizens based on their political speech, and whether the courthouse doors can be closed to their constitutional claims. It’s the modern descendant of the André question—how much process the government owes those it deems a threat—and the Supreme Court is now being asked to answer it.

Trump administration asks US Supreme Court to allow detention of pro-Palestinian scholar | Reuters · Al Jazeera · US News

Now a sobering story about accountability within the judiciary itself: a federal judge in Michigan will take a one-year leave from the bench to continue treatment after a drunk-driving conviction. U.S. District Judge Thomas Ludington, who is 72, was arrested in October 2025 in Emmet County, in northern Michigan, with a blood-alcohol content the prosecutor put at 0.270—more than three times the legal limit, a level that falls under Michigan’s “super drunk” enhanced-penalty statute. He pleaded no contest in April to misdemeanor operating while intoxicated, was sentenced in May to six months of probation and over $1,000 in fines and costs, and is now stepping back from his caseload for a year to undergo treatment, counseling, and testing. Let me talk about why this matters legally, because it’s not about the criminal case, which is resolved—it’s about judicial accountability. Federal judges have life tenure under Article III; they can only be removed by impeachment, which is a deliberate constitutional design to protect judicial independence. That means when a federal judge has a personal crisis like this, there’s no boss to suspend them—the system relies heavily on self-regulation and on the judicial-conduct machinery, where a judge’s colleagues, through the circuit’s judicial council, can investigate and impose measures short of removal. Ludington’s decision to take a voluntary leave and pursue treatment is that self-regulation working roughly as intended: an acknowledgment that a judge who decides other people’s cases, including drunk-driving and addiction cases, has to hold himself to the standard he applies to others. And there’s a compassionate dimension here too, one that connects to a story we covered recently about the legal profession’s mental-health and substance-use crisis—that a 0.270 BAC is not “having a few too many,” it’s a sign of a serious disease that the profession is finally learning to treat as illness rather than mere misconduct. The significance is a small but real illustration of how accountability functions for the most independent actors in our legal system, and a reminder that the bench is not immune to the profession’s struggles.

Michigan federal judge to remain on leave, undergo treatment after drunk-driving arrest | Reuters · The Detroit News · UpNorthLive

And finally, DoorDash has reached a $131.5 million settlement with New York City over how it paid—or underpaid—its delivery workers. The city found that DoorDash either underpaid workers or took too long to pay them, and the settlement breaks down in a revealing way. Nearly $115 million goes as relief to about 264,000 workers—including roughly $83 million to resolve a fight over how to calculate pay for workers who were logged into the app and on-call but not actively making a delivery, plus $12.3 million for payments that were missed or arrived days or weeks late—and there’s a $16.7 million fine on top. The legal backdrop is the gig-economy pay wars, and New York City has been the most aggressive jurisdiction in the country here. The city enacted a first-in-the-nation minimum-pay standard for app delivery workers, and this settlement enforces the promise behind it—that the time a worker spends available and waiting for orders is compensable, not free labor the platform gets to ignore. That “on-call time counts” principle echoes classic wage-and-hour law, the same kind of question courts have long wrestled with under the Fair Labor Standards Act about when waiting time is working time. There’s also a striking allegation in the background: the city accused DoorDash and Uber Eats of engineering “design tricks” to deprive workers of more than $550 million in tips—a reminder that in the gig economy, the design of the app itself can be the mechanism of wage theft. Now, one honest caveat for workers: DoorDash’s median payout works out to about $48 per worker, so while the headline number is large, the individual relief is modest. The significance is that this is a major enforcement win for a city that has led the nation in regulating gig work, and further confirmation that “flexibility” and app-based independence don’t exempt companies from the basic legal obligation to pay people for their time.

DoorDash reaches $131.5 million settlement with NYC over delivery workers’ pay | Reuters · NBC News · Bloomberg

This Day in Legal History: The Preliminary Emancipation Proclamation

On September 22, 1862, following the Union’s costly victory at Antietam, President Abraham Lincoln issued the preliminary Emancipation Proclamation. It gave the Confederate states an ultimatum: return to the Union by January 1, 1863, or the enslaved people in the rebelling states “shall be then, thenceforward, and forever free.” The Confederacy did not yield, and on New Year’s Day 1863, Lincoln issued the final Proclamation, recasting the Civil War as a war against slavery and setting more than three million people on the path to freedom.

For legal history, the Proclamation is a landmark study in the scope—and the limits—of executive power. Lincoln did not act under a statute passed by Congress; he acted as Commander-in-Chief, framing emancipation as a “fit and necessary war measure” to weaken the Confederacy. That’s why the Proclamation, by its own terms, reached only the areas in rebellion—not the border states loyal to the Union, where Lincoln doubted his war powers extended. It was a document acutely aware of its own legal boundaries. Lincoln understood that a wartime executive order might not survive the war’s end, which is precisely why he pushed so hard for the 13th Amendment—to place the abolition of slavery on the permanent, unshakable footing of the Constitution rather than the contested ground of a presidential proclamation.

The significance of September 22, 1862 is that it is one of the most consequential exercises of executive authority in American history—and a permanent lesson in that authority’s nature. It showed both the enormous power a president can wield in a crisis and the reason such power is inherently limited and temporary, requiring the other branches to make it durable. That tension—between decisive executive action and the checks that legitimate and constrain it—runs directly into my column today, which is, at its heart, about exactly that balance.

Paramount has settled with California and eleven other states, clearing one of the biggest remaining hurdles to its roughly $110 billion merger with Warner Bros. Discovery—a deal that would reshape Hollywood. Let me set the stage. Back in July, a coalition of twelve states, led by California Attorney General Rob Bonta, sued to block the merger, arguing in their complaint that combining Paramount and Warner Bros. would “extinguish competition” in the entertainment industry. And you can see why they worried: this deal would unite two of Hollywood’s biggest film studios, two major streaming services, and—critically for the news business—two of the largest cable-news operations, CBS and CNN, under a single owner. This is state antitrust enforcement, and it’s a reminder that state AGs are increasingly aggressive independent players in merger review, not just spectators to the federal agencies. The settlement, which came together over the weekend after four holdout states came around, includes some genuinely interesting concessions. First, an output commitment: Paramount pledges to release 30 films a year for the first two years and 32 a year for the next three, or pay a penalty—a remedy aimed squarely at the fear that the merged giant would slash output and starve theaters. But the most striking term, from a media-law perspective, is this: the states secured a commitment to independent editorial boards for CBS and CNN. Think about what that addresses—the concern that concentrating this much news media under one owner threatens editorial independence, that the danger of media mergers isn’t just higher prices but a narrowing of the free press. The significance is twofold: it clears a major path toward closing one of the largest media mergers in history, and it shows antitrust remedies stretching beyond the usual price-and-output concerns into the territory of safeguarding editorial independence—a novel and telling wrinkle in an age of consolidated media.

Paramount settles with California, other states, clearing major hurdle for Warner Bros | Reuters · Washington Post · NBC News

Now a story that will resonate with anyone who’s ever sat for the bar: the vendor behind California’s disastrous 2025 bar exam has agreed to a class-action settlement. Longtime listeners may recall the debacle—the February 2025 California bar exam was marred by serious technical failures, with test-takers reporting crashing software, login problems, and lost answers on the online platform, in what became a genuine crisis for the people whose careers hung on that test. The company that administered it, ProctorU, doing business as Meazure Learning, has now agreed to settle the examinees’ proposed class action for about $1.35 million. Under the deal—which still needs approval from a federal judge in the Northern District of California—roughly 4,100 examinees would get full refunds of the $153 laptop fee they paid to take the exam. One quick clarification, because you may see a bigger number floating around: the class-action figure is $1.35 million, but separately, the State Bar of California itself sued the vendor and secured a $5.25 million settlement—so the total the vendor is paying across both actions is meaningfully larger than the class-action number alone. The legal framing is a straightforward but important one: this is a consumer/contract and negligence theory—the plaintiffs alleged the company deployed malfunctioning software despite knowing about glitches weeks in advance. That last part, the alleged prior knowledge, is what elevates it from unfortunate technical failure toward actionable misconduct. The significance goes beyond the dollars, which are modest—a $153 refund doesn’t remotely capture the stress and career disruption of a botched bar exam. It’s a cautionary tale about the high-stakes migration of critical, gatekeeping exams onto proprietary software, and about accountability when that technology fails the people who depend on it. As more of the legal profession’s own infrastructure goes digital, the reliability of the vendors behind it becomes a real professional-responsibility concern.

California bar exam software provider to pay $1.35 million in class action over test | Reuters · Bloomberg Law · Law360

And finally, in my column for Bloomberg Tax this week, I take on the messy afterlife of the tariff wars—specifically, what happens when tariffs get unwound. Billions in tariff refunds are now flowing back to U.S. companies after the Supreme Court struck down the IEEPA tariffs earlier this year, and my argument is that the whole refund process reveals a deep design flaw: the government is good at returning cash to the businesses that legally paid it, but it has no way to get that money back to the consumers who actually bore much of the cost. Here’s the core problem. Tariffs legally fall on the importer—the company that writes the check—but economically, that burden gets passed down the supply chain into higher prices for distributors, retailers, and ultimately consumers. So when the tariff is refunded, the money goes back to the importer, not to the people who really paid. And companies do whatever they want with that windfall—pay down debt, reward workers, or lower prices on something totally unrelated. My favorite illustration in the piece: imagine you overpaid for a coffee machine last year because of the tariff, and the company uses its refund to discount patio furniture this year. The patio-furniture buyer gets a subsidy funded by your coffee-machine overpayment. That’s only a “refund” if you treat all consumers as one undifferentiated blob rather than actual individuals. Now here’s the legal heart of my argument, and it draws on the tax code. Section 6416 of the Internal Revenue Code already solves a version of this for federal excise taxes: a business generally can’t get a refund just because it remitted the tax—it has to show it either didn’t pass the tax on to customers, or it repaid them, or they consented. Tariffs have no comparable mechanism. So my proposal is that Congress should require any temporary tariff to contain an unwinding mechanism from the very beginning—specifying who gets refunded, what happens when the burden was shifted downstream, whether claims accrue interest, and how Treasury should account for potential refund liability while the tariff is even in effect. And this ties directly to today’s legal-history theme: the Supreme Court, in striking down those tariffs, essentially treated tariffs as a branch of the taxing power. So maybe, I argue, we should start treating their unwinding with the same seriousness—and the same built-in checks—we give other taxes. Given that this administration already exceeded the tariff authority Congress delegated it, this isn’t a hypothetical worth shrugging at. It’s a design problem Congress should fix before the next tariff, not after.

Tariffs Need a Checks-and-Balances System From the Beginning | Bloomberg Tax