We’ve launched Minimum Competence CLE, and our first course is now available completely free. Researching Federal Tax Issues After Loper Brightlooks at how the Supreme Court’s decision ending Chevron deference changes the way lawyers should research and evaluate Treasury regulations, IRS guidance, and other federal tax authorities. Take the course and earn CLE credit at cle.minimumcomp.com. This Day in Legal History: The National Park Service Is Created On August 25, 1916, President Woodrow Wilson signed the National Park Service Organic Act, creating a new federal bureau inside the Department of the Interior to manage the country’s national parks and monuments. Before this, the roughly 35 parks and monuments the government had accumulated were administered in a haphazard, piecemeal way—some by the Army, some by scattered Interior officials—with no unified mission and no consistent protection. The genius of the Organic Act is compressed into a single, oft-quoted sentence that has generated a century of law and litigation. It directed the new Park Service to “conserve the scenery and the natural and historic objects and the wild life therein” and to “leave them unimpaired for the enjoyment of future generations.” That phrasing contains a built-in tension that lawyers and the courts have wrestled with ever since: the Park Service is told both to provide for the public’s enjoyment of the parks and to leave them unimpaired—two goals that can pull hard against each other when you’re deciding whether to build a road, permit a concession, or limit the number of visitors. That “dual mandate” is the central puzzle of public-lands law. The significance of August 25, 1916 is that it enshrined in law a genuinely radical idea: that some places belong to the whole public, in perpetuity, held in trust by the government for people who haven’t been born yet. Today the Park Service manages more than 400 sites across all fifty states and welcomes hundreds of millions of visitors a year. The writer Wallace Stegner, who gave us today’s opening line, called the national parks “the best idea we ever had”—absolutely American, absolutely democratic. On a day whose news is heavy with contested executive power, it’s a nice reminder that the law is also the instrument through which a society decides to preserve something for the common good. The Supreme Court has handed the Trump administration an interim victory on mail-in voting—but it’s important to be precise about what the Court did and didn’t do. In a 6-3 decision along ideological lines, the justices lifted an injunction that a Massachusetts federal judge imposed in June, which had blocked key parts of Trump’s executive order restricting mail-in voting in California and 22 other states that sued. That order aims to create a federal list of eligible voters and, controversially, to have the Postal Service deliver ballots only to people on that list. Here’s the crucial nuance. This is a stay, not a ruling on the merits—the Court did not decide whether the president actually has the authority to reshape how states run their elections. It merely paused the lower court’s block while the litigation continues. And it’s partial: the Postal Service’s role in the scheme remains blocked nationwide. Justice Sotomayor, in dissent, made the point sharply—she wrote that the decision doesn’t suggest the executive branch has any constitutional or statutory authority to do this, it just postpones the day of reckoning. This is the shadow docket in action, a theme we keep returning to: the Court altering the rules for a national election months before that election, through an emergency order, without a full merits decision or a written explanation of its reasoning. The significance is that, for now, parts of Trump’s order can take effect for the November midterms—a real-world consequence—even as the fundamental question of whether any of it is lawful remains formally undecided. The status quo the courts had maintained all summer just shifted, at least temporarily, and it shifted on the emergency docket. US Supreme Court lifts judicial hurdle to Trump’s mail-in ballot curbs | Reuters Washington Post · NPR The administration is preparing what would be the largest mass visa revocation in U.S. history: according to the Associated Press, the State Department plans to revoke the tourist and business visas of up to 200,000 foreigners who have applied for or are seeking asylum. The targets are holders of so-called B1 and B2 visas—the standard business and tourism visas—issued between 2016 and 2026, whose holders later sought asylum, with the action coordinated between the State Department and the Department of Homeland Security. The legal logic is worth understanding. A visitor visa rests on a premise: that you intend to come temporarily and then go home. When someone on a tourist visa applies for asylum, the government’s position is that they’ve revealed an intent inconsistent with that temporary-visitor status—so it’s moving to strip the visitor visa. Importantly, officials say this wouldn’t necessarily mean immediate deportation; most people with pending asylum cases would be recategorized, losing their visitor status but not instantly removed. Still, the scale is staggering and unprecedented, and it raises hard questions. Seeking asylum is a legal right, protected by statute and international obligation, and there’s a real tension in penalizing people’s immigration status precisely because they used the lawful asylum process. There are also due-process concerns lurking in any mass action that recategorizes 200,000 people, and a potential chilling effect on those weighing whether to come forward and claim protection. The significance is that this fits the aggressive immigration-enforcement pattern we’ve tracked all summer—but at a scale that’s genuinely new, using visa revocation as a lever against the asylum system itself. US plans to revoke up to 200,000 tourism, business visas, AP reports | ReutersWTOP · Houston Public Media / NPR And a story close to home for everyone in this profession: Google has launched a version of its Gemini AI built specifically for lawyers. Google Cloud unveiled “Gemini Enterprise for Legal,” an agentic AI platform purpose-built for legal workflows—with specialized skills, connectors to legal research and document systems, and an ecosystem of partner tools. And its launch customers are not fringe experimenters; they’re elite firms: Cleary Gottlieb, Freshfields, Weil, and Williams & Connolly. When firms of that caliber sign on as launch partners, it’s a signal that AI in legal practice has moved from novelty to infrastructure. Here’s the significance, and the tension. On one hand, this is the mainstreaming of AI in law—the same shift that, as we discussed a few weeks ago, is thinning out entry-level hiring at big firms as software absorbs the document review and first-draft work once done by junior associates. Google is now competing directly with the specialized legal-AI companies for that market, and its entry accelerates everything. On the other hand, this collides with a problem we’ve covered again and again: the fake, hallucinated citations, the AI-generated errors, even the litigant hiding prompt injections in his filings. A more powerful, more deeply integrated AI tool doesn’t dissolve a lawyer’s duties of competence, confidentiality, and candor—it raises the stakes on them. The significance is that the profession is being reshaped in real time. The tools are getting better and more embedded, the biggest firms are adopting them, and the hard questions—about verification, about who’s accountable when the machine is wrong, about what happens to the training pipeline for young lawyers—are all arriving at once. Google expands Gemini AI platform for law firms, lawyers | ReutersArtificial Lawyer · Google Cloud And finally, in my column for Bloomberg Tax this week, I dig into a big loss for Maryland—and a valuable lesson for every other state eyeing the digital economy for revenue. Maryland’s first-in-the-nation digital advertising tax just took a potentially fatal blow, with the state tax court siding with Google, Apple, and Peacock in their challenges. My argument is that this doesn’t prove you can’t tax digital advertising—it proves Maryland went about it the wrong way, by designing a tax around the companies it wanted to pay rather than the activity it wanted to tax. Two design choices doomed it. First, Maryland taxed digital advertising while leaving comparable non-digital advertising—billboards, print, radio, television—completely alone. Second, the tax only kicks in for companies above a huge global revenue threshold, and the rate actually climbs, from 2.5% up to 10%, based on a company’s worldwide revenue, not its Maryland activity. The court saw that for what it was: a structure that, in practice, hits big out-of-state tech companies while sparing basically every in-state business, which runs into both the dormant Commerce Clause and the federal Internet Tax Freedom Act. As I put it, using worldwide corporate revenue to set the rate on identical in-state transactions is just a backward way of making big companies pay more—two companies doing the exact same amount of business in Maryland can end up with wildly different bills. There’s also a cautionary detail I love, because it says everything: Maryland passed this tax in 2021, but by 2025 the state comptroller had to hire an outside expert just to figure out what “digital advertising services” even means—introducing terms like “programmatic” and “visual” that weren’t in the statute at all. If you need to hire an expert four years after enactment to ascertain what your own tax taxes, maybe the problem isn’t that taxpayers are being difficult. And the fisca