This Day in Legal History: The Weimar Constitution On August 11, 1919, President Friedrich Ebert signed the Weimar Constitution into force, creating the first parliamentary democracy in German history. It came into effect a few days later, ending the provisional government that had followed Germany’s defeat in the First World War and the fall of the monarchy. On paper, it was a strikingly modern and liberal document—and its story is one of the most important cautionary tales in all of constitutional law. The Weimar Constitution rested on genuinely progressive foundations: popular sovereignty, a separation of powers, and an extensive bill of fundamental rights, including—remarkably for 1919—formal equality between men and women and the right to vote for both. It established a directly elected president, a parliament, the Reichstag, and a framework that looked, in many respects, like a model twentieth-century democracy. For a document drafted in the chaos of postwar Germany, it was an ambitious attempt to build a rights-respecting republic from the ground up. But the Weimar Constitution is remembered today largely for the flaw that helped destroy it: Article 48, which gave the president sweeping power to rule by emergency decree, suspending civil liberties when he deemed public order to be threatened. That emergency clause was invoked more and more frequently as the republic’s politics fractured—and in 1933 it became the legal mechanism through which the Nazi regime dismantled democracy from the inside, using the constitution’s own emergency powers to suspend rights and seize dictatorial control. Which brings us back to today’s opening quote: the Weimar experience is the definitive real-world example of Justice Jackson’s warning that an emergency power, once available, “lies about like a loaded weapon, ready for the hand of any authority that can bring forward a plausible claim of an urgent need.” The significance of August 11, 1919 is that it stands as an enduring lesson that a constitution’s protections are only as durable as the limits it places on emergency power—that a democracy can be subverted not in spite of its own laws, but through them. Luigi Mangione is due back in court for a pretrial hearing ahead of his murder trial, set to begin September 8, in the December 2024 killing of UnitedHealthcare CEO Brian Thompson outside a Manhattan hotel. Mangione, who is 28, has pleaded not guilty to murder, weapons, and forgery charges brought by the Manhattan District Attorney, and he has separately pleaded not guilty to federal stalking charges. At today’s hearing before Justice Gregory Carro, the parties are expected to take up jury selection and the thorny question of public access to what is shaping up to be one of the most closely watched trials in years. Let’s talk about the real legal challenges here, because this case is unusual on several fronts. First, Mangione faces parallel state and federal prosecutions arising from the same killing—something the Constitution permits under the “dual sovereignty” doctrine, even though it can feel like being tried twice. Second, the defense has signaled it may argue that Mangione lost control of his actions due to an extreme mental-health breakdown, and notably has not conceded that he was even the shooter, despite prosecutors citing surveillance video, DNA, ballistics, and a notebook. And third—the elephant in the courtroom—is that this case became a cultural flashpoint, with a wave of public anger at the health-insurance industry curdling, in some corners, into sympathy for the accused. That makes jury selection genuinely difficult: finding impartial jurors amid saturation coverage and strong public feeling is exactly the kind of problem that tests the machinery of a fair trial. The significance is that beyond the headlines, this is a serious test of whether the system can give a fair, dispassionate trial to a defendant the public has turned into a symbol—which is precisely when the guarantees of criminal procedure matter most. Luigi Mangione due in court ahead of insurance CEO murder trial | ReutersUPI · 6abc President Trump has said that whether to revive the $1.8 billion “anti-weaponization fund” is now up to his newly confirmed attorney general, Todd Blanche—a statement that lands like the punchline to a story we’ve been telling for weeks. Recall the sequence: Blanche rescinded the fund on the eve of his confirmation to win over Republican holdouts, but Trump then acknowledged that he himself never signed the rescission order, which left the door open. Now, with Blanche confirmed as attorney general by that razor-thin 50-49 vote, Trump is saying the future of the fund rests with Blanche. And that is exactly the scenario that critics warned about. Think back to what senators extracted as the price of confirmation: a promise to kill a fund that would have used taxpayer money to pay people, including January 6 defendants, who claimed to be victims of the prior Justice Department. But a promise made by the executive branch can be unmade by the executive branch, and here the president is openly signaling that his former personal lawyer, now running the Justice Department, could bring it back. The significance goes to the very heart of the concern that dogged Blanche’s nomination: the independence of the attorney general. When the president publicly frames the revival of a controversial payout fund as a decision for “his” attorney general to make, it underscores the worry that the Justice Department’s most consequential choices may run through the lens of the president’s personal and political interests. The confirmation fight is over, but the question it raised is very much alive. Trump says future weaponization fund is up to Blanche | ReutersForbes · Axios A federal appeals court has cleared thousands of lawsuits to proceed against the major social-media companies over claims they deliberately designed their platforms to addict young users. The San Francisco-based Ninth Circuit rejected the bid by Meta, Google’s YouTube, TikTok, and Snapchat to escape roughly 2,400 lawsuits consolidated in federal court, brought by states, municipalities, school districts, and individuals. The legal fight here is all about Section 230 of the Communications Decency Act—the 1996 law that shields online platforms from liability for content their users post. The companies argued that Section 230 also immunizes them from claims that they failed to warn about, or engineered, the addictive nature of their platforms. The court said no, and the distinction it drew is the crucial one: Section 230 protects a platform from being sued over what its users say, but it does not necessarily protect the platform’s own product-design choices—the infinite scroll, the algorithmic feeds, the notification schedules engineered to maximize engagement. Claims aimed at that design, the court reasoned, are different from claims aimed at user content. If you’ve been listening, you know this is the federal counterpart to everything we’ve tracked at the state level—the New Mexico judgment, the Tennessee trial, the multistate penalty fight. The significance is that the industry’s most powerful legal shield, Section 230, is being narrowed: plaintiffs are increasingly framing their cases around addictive design rather than harmful content, and courts are increasingly letting those cases through. That reframing may prove to be the key that unlocks the courthouse door for the whole wave of social-media harm litigation. US appeals court allows thousands of lawsuits against social media companies to proceed | ReutersNBC News · Al Jazeera And finally, in my column for Bloomberg Tax this week, I dig into a question hiding underneath the AI gold rush: who’s actually paying for it? My starting point is a striking data point—Microsoft’s current federal tax expense fell year over year from about $14.1 billion to just $2.5 billion, even as its revenue surged. And a big part of that drop comes from accelerated deductions in last year’s massive 2025 tax law, which restored 100% bonus depreciation—letting companies write off the full cost of huge investments, like AI data centers, in the very first year. My core argument is that the tax law is financing the AI infrastructure boom, but it isn’t creating it—and that distinction matters enormously before anyone declares the policy a success. The political narrative writes itself: capital spending soared after the tax cut, therefore the tax cut worked. But I think correlation is doing an awful lot of heavy lifting there. Microsoft and its rivals aren’t pouring tens of billions into AI infrastructure because Congress rediscovered depreciation. They’re doing it because falling behind in AI could threaten their core businesses—it’s become about as close to a strategic necessity as it gets. And here’s the tell: Microsoft announced its roughly $80 billion AI data-center plan in a blog post in January 2025, a full six months before the tax law was even signed. The trajectory was public, budgeted, and well underway before the depreciation rules changed. So my point to Congress is: prove it. Before congratulating yourselves, you owe the public evidence that these tax benefits actually changed corporate behavior rather than just rewarding investments companies were already racing to make. And I want to push back on the idea that this is “just” a timing benefit and therefore basically free. Timing isn’t free—letting a company keep its cash now in exchange for tax revenue later is like giving it an interest-free loan, and every dollar of accelerated depreciation is a dollar the Treasury can’t use for something else this year. That’s a real opportunity cost, felt in deficits or forgone public investment. My recommendation is concrete: Congress should require the Treasury and the J
