From Thursday's Schulte v. Leners, decided by Judge Frank Easterbrook, joined by Judges Doris Pryor and Joshua Kolar:
Elizabeth Schulte co-owns real property in Cedar Lake. The Town is governed by a three-member Board of Supervisors, of which Kenneth Leners served as chairman in September 2022. The Town maintains a website (townofcedarlake.org), which it uses to post things such as agendas, officials' contact information, and updates on pending business. According to Schulte's complaint, which we take as true at this stage, until September 2022 the website allowed members of the public to comment on the Town's posts. Members of the public often expressed their thoughts and opinions on the original post and discussed other issues. The Town did not limit who could comment or what they could say.
On September 13, 2022, Chairman Leners, who possessed editorial control of the website, posted a notice titled "Update on Recreational Residential Dwellings." The notice criticized the use of recreational vehicles in Cedar Lake and advocated greater regulation of RVs. Schulte—who owns an RV—disagreed and said so in the comment section. Others joined. Fifteen public comments addressed the substance of Leners's post; thirteen took issue with it.
A week later Leners posted a reply to Schulte, calling her message "disrespectful" and announcing that he'd "no longer allow [her] to comment on this site." Schulte answered in kind, disputing both Leners's views about RVs and his characterization of her posts. Leners promptly deleted her reply. He did the same to at least two other critics in the comment thread: label their comment disrespectful, announce a ban, and delete their posts.
On September 22, after threats of legal action, Leners removed all remaining comments from his post, again citing "disrespect[]." At an October 17 town meeting, Leners sought to justify his actions as a "judgment call" based on what he viewed as "derogatory comment[s]." The Board eventually voted to end comment sections altogether. The deleted comments from Leners's post were later restored to the site, but in a standalone thread unconnected to the original post. As of the date of Schulte's amended complaint, Leners's post of September 13 remained on the website without comments….
Schulte sued, claiming Leners' "deleting individual comments" and "deleting all comments under his post" violated her First Amendment right, and the court allowed the case to go forward:
Northern District of Texas decides that registration requirements for non-taxed NFA firearms are no longer grounded in Congress’ power to lay and collect taxes.
In a decision issued on August 5, U.S. District Judge James Wesley Hendrix (N.D. Tex.) ruled in Silencer Shop Foundation v. Bureau of ATF that the One Big Beautiful Bill Act, which repealed the taxes for the making and transferring of certain firearms under the National Firearms Act (NFA), left the law without any basis for its registration requirements in the authority of Congress to "lay and collect Taxes." U.S. Const. art. I, § 8. cl. 1. The firearms affected are short-barreled rifles, short-barreled shotguns, silencers, and "any other weapon[s]," all defined in 26 U.S.C. § 5845. Machineguns and destructive devices remain subject to the $200 tax.
This result is a no-brainer, as Congress enacted the NFA as a tax measure under the Internal Revenue Code and the Supreme Court upheld it as such in Sonzinsky v. United States (1937), which the Court reaffirmed in NFIB v. Sebelius (2012), the Obama-care case. Citing this author's article on the subject, the district court wrote that "the NFA was Congress's attempt to regulate so-called 'gangster weapons' linked to Prohibition Era violence." The purpose of the NFA requirement that the "firearms" under the Act be registered was to keep track of who paid the tax. Repeal of the tax pulled the rug out from under the registration requirement.
Silencer Shop next considered the government's argument that the regulatory requirements for the making and transfer taxes support the collection of the NFA's special occupational taxes (SOTs), which persons in the business of manufacturing, importing, and dealing in NFA firearms must pay annually. But collection of those taxes is implemented through the registration requirements for such businesses and the penalties for noncompliance. The regulatory scheme for the SOTs is completely distinct from that for the making and transfer taxes.
Quoting Printz v. United States (1997), the court next turned to the government's claim that the provisions at issue are valid under "the Necessary and Proper Clause—the 'last, best hope of those who defend ultra vires congressional action.'" But the making and transfer regulatory provisions are not "necessary" to administration of the regulatory scheme for the SOTs. Nor are they "proper," as Congress has no plenary police power such as the States have.
ATF claimed that that the challenged registration provisions support the collection and enforcement of the SOTs. But registration information is barred from use in criminal prosecutions per 26 U.S.C. § 5848(a) and the Fifth Amendment privilege against self-incrimination, see United States v. Freed (1971).
The government argued next that the NFA provisions may be sustained under the Commerce Clause. While courts respect the lawmaking power of Congress, the court wrote, that is "not a get-out-of-unconstitutionality-free card." By its plain text, the NFA is a taxing statute. Nothing in the provisions at issue mention interstate or foreign commerce. By contrast, Title I of the Gun Control Act (GCA) includes such commerce as the jurisdictional hook of every one of its provisions. That is, with one exception.
That exception is the ban on possession of a post-1986 machinegun, 18 U.S.C. § 922(o). The Fifth Circuit had suggested in dicta in United States v. Ardoin (1994) that, although ATF would not collect NFA taxes for new machineguns because of § 922(o), that part of the NFA as applied might be sustained under the Commerce Clause. But as the Silencer Shop court points out, Ardion preceded the Supreme Court's decisions in Lopez and Morrison, which invalidated two federal laws respectively – the Gun Free Schools Act and the Violence Against Women Act – as police power measures having no basis in the Commerce Clause.
That left only the Second Amendment claim, but at oral argument, all of the parties agreed that that issue need not be decided if the court found that the NFA provisions exceed the powers of Congress. That claim was thus forfeited. I previously addressed the arguments in the briefs here.
With that, the court issued a permanent injunction as the final judgment enjoining the Department of Justice and ATF from "from enforcing the challenged NFA provisions as to untaxed firearms against the plaintiffs and, where applicable, the plaintiffs' agencies, political subdivisions, members, and customers—both current and future." The plaintiffs include individuals, nonprofit associations, businesses that deal in NFA firearms, together with Texas and thirteen other states. The injunction's scope is broad as to the businesses inasmuch as they may sell firearms to both current and future customers. Its scope is narrow as applied to the fourteen state plaintiffs, who are not subject to the NFA requirements when the state itself and its political subdivisions acquire NFA firearms, but whose citizens are not covered by the injunction.
The affected firearms in Silencer Shop include short-barreled rifles, short-barreled shotguns, silencers, and "any other weapon[s]," AOWs. The case was consolidated with Jensen v. ATF and the court's holding mostly applies to both cases equally, except that the Jensen plaintiffs raised no AOW claim, and so the court's injunction doesn't apply to those plaintiffs as to AOWs.
This is not a universal, nationwide injunction, which is precluded by the Supreme Court's decision in Trump v. CASA (2025). Moreover, imposition of a universal injunction would short-circuit review of similar challenges pending in the Eastern District of Missouri and the Eastern District of Kentucky.
The court stayed the injunction order for seven days to give the government an opportunity to appeal to the Fifth Circuit, which seems sure to issue its own stay pending appeal. The challenges could easily end up in the Supreme Court. Given the longstanding, solid anchoring of the NFA in the tax power, plaintiffs have reason to be optimistic about ultimately prevailing. But that doesn't mean the sky will fall. Recall that the Senate had been poised to delete the affected firearms from the NFA altogether, but the parliamentarian ruled that to be disallowed by the Byrd Rule concerning reconciliation bills. That's why the Senate simply zeroed out the tax. See my post here.
As Winston Churchill said after the British victory at the Battle of El Alamein, "This is not the end. It is not even the beginning of the end. But it is perhaps the end of the beginning."
I appreciate the emails I received from my twoposts on starting a Substack. I am also grateful to David Lat who chatted with me at some length about the pros and cons of Substack. David has one of the most successful legal substacks, so he knows the business well.
I'll put my tentatively final thoughts up front: starting a paid Substack probably isn't worth it.
First, David explained how conversion rates actually works. In short, almost all substacks offer different tiers of subscriptions. Everyone can sign up for a free subscription that grants access to some content. Others sign up for a paid subscription that grants access to all content. The percentage of unpaid subscriptions that become paid is known as the conversion rate. I've seen some estimates that the Substack conversion rates hover around 10%. For example, if you have 1,000 total subscribers, about 100 of them are paid. David relays that the average conversion rate is probably closer to 5%. With the same number of unpaid subscribers, there may be about 50 paid subscribers.
Second, subscriber numbers are not static. People who cross the threshold to pay are very willing to cancel the subscription. If there is not enough content, they may think the price is not worth it, and unsubscribe. If there is too much content, and their inbox is being flooded, they may also unsubscribe. There is also a cyclical nature where people tire of a particular site and seek out other subscriptions. I've talked to several Substackers, and they all indicate that subscription numbers have largely plateaued. Once the market reaches its saturation rate, new subscribers are hard to find. Plus, old substacks had something of a first mover advantage. New sites will struggle. We may have already reached peak Substack.
Third, much work goes into keeping the conversion rate high and the unsubscribe rate low. Authors have to give considerable thought into how much content to generate, what type of content to generate, when to publish the content, and how to make the emails worth opening. If the author fails at any or all of these tasks, the subscriber rate will go down. Authors will necessarily obsess over these matters, and I think this concern will necessarily impact the quality and quantity of materials produced. Perhaps some authors have figured out the right balance, but many do not. Some substacks will simply fade away in relevance, and readers may never know.
These three points above have impacted how much I could reasonably expect to make. Assuming a $5/monthly subscription, to make $1,000 per month, I would need 200 paid subscribers and about 4,000 total subscribers. To make $10,000 per month, I would need 2,000 paid subscribers and about 40,000 total subscribers. Based on my rudimentary calculations, I imagine my paid subscriber rate at the outset would be towards the bottom of this range. Maybe it would grow over time, but it would inevitably plateau as I saturate the market. As one person told me by email, how many posts can I write about Chief Justice Roberts's political approach to judging? (Hint: another one is coming tomorrow.)
In candor, a few thousand dollars each month would be some nice pocket change to have, but not enough to justify the change. Moreover, I enjoy the current routine: I blog when I want, don't blog when I'm doing other things, and do not much care how many people actually read the post.
Anyway, these are my tentatively final thoughts for now. I of course remain open to revisiting this topic in the future. Thanks to all who contacted me!
A detailed Marquette Law School poll finds public views of the Court are modestly negative, but most major recent decisions enjoy majority public support.
Back in May, I wrote about survey data indicating a modest boost in the Supreme Court's popularity with the public, possibly caused by its February decision in the tariff case (which I helped litigate). Since then, the Supreme Court has issued its major decisions in a large number of cases, including those on birthright citizenship, the Second Amendment, executive removal power, and more. Recently, Marquette Law School came out with a new poll assessing public perceptions of the Court in the wake of those rulings. It finds that the Court as a whole remains modestly unpopular (47% approve, 53% disapprove), but that most of its major recent decisions have majority public support. The Court also remains a bit more popular than just before the tariff ruling, and much more popular than the other two branches of government. There are other interesting findings, as well, including on perceptions of individual justices (about which most of the public is generally ignorant).
As discussed below, the Marquette survey has a few unfortunate methodological problems. But it nonetheless provides some useful information about public perceptions of the Court.
Before going on, I will emphasize, yet again, that the Court's popularity (or lack thereof) says little about the quality of its decisions. Survey data consistently shows that most Americans know very little about the Constitution and the Court's work, and a majority of the public cannot even name one Supreme Court justice and know little about them (the Marquette poll further confirms this latter point). I also do not claim that majority public support for decisions I like (including the tariff ruling!) "proves" those decisions are correct. It does not. But, for reasons to be discussed below, public opinion about the Court does matter, in so far as it is important to protecting the institution of judicial review against potential attacks from both the right and the left. It may also be of interest to students of the Court and students of public opinion (I happen to be both).
The Court's 47% approval rating is slightly higher than the 44% it got in the same poll in mid to late 2025, and about the same as the 46% it got in May (after the tariff case, but before the other rulings). An April Marquette poll showed 42% approval. This is consistent with other survey data indicating a slight improvement in early 2026, possibly driven by the tariff decision. Though the Marquette poll shows more stability than some of the others, and fluctuations between 42 and 47 could just be due to random variation.
The Court also remains much more popular (or, rather, much less unpopular) than the other two branches of government. The Marquette survey finds that President Trump has only a 38% favorability rating (a whopping 60% view him unfavorably), and both congressional Republicans and congressional Democrats, are also highly unpopular, probably even more so than Trump. The same goes for the Republican and Democratic parties generally. This, too, is consistent with previous polls showing that public views of the Court are more favorable than those of the other two branches of government.
Marquette also surveyed public attitudes on 13 significant decisions made by the Court within the last year. And all but two command majority support, including many supported by large majorities:
Marquette University
The tariff case (66% support), allowing states to ban transgender athletes from women's sports (65%), and Trump v. Cook - the decision protecting the autonomy of the Federal Reserve Board (70%) are all highly popular. Louisiana v. Callais - the Voting Rights Act decision that attracted strong left-wing ire - enjoys modest majority support (53%). The only two decisions surveyed that don't have majority support are Trump v. Slaughter (ruling that the president can remove the heads of most other independent agencies), and a ruling allowing parties to coordinate campaign expenditures with campaigns. And even those two decisions are only modestly unpopular, getting support from 49% of the public, and 47% respectively.
There are a few methodological issues with this part of the Marquette survey. The survey's descriptions of the rulings in the questions are fairly detailed and precise. But that very fact might make them hard to understand for respondents who don't know much about the issues involved (which is likely to be a large proportion of them). In addition, all of these questions are structured as "forced choice" questions, on which respondents do not have the option of answering "don't know" or "no opinion." That means the survey is capturing a lot of responses by people who don't actually have meaningful opinions on the case in question.
In addition, the Court included one case it probably should not have, and omitted at least one that should have been included. The Court did not really rule on E. Jean Carroll's suit against Trump. As the Marquette survey notes, it merely refused to review the lower court decision in favor of Carroll. This is not actually a precedent-setting decision by the Court, and not that significant for anyone but Trump and Carroll. By contrast, the survey wrongly omits the Court's terrible decision in the Haitian TPS case, which will both have a massive immediate impact on hundreds of thousands of people and set a bad precedent for future executive immigration policies motivated by racial and ethnic prejudice.
Despite these caveats, it is clear that most of the Court's recent major decisions enjoy majority public support, some overwhelmingly so. And none are highly unpopular.
Marquette also surveyed public opinion about individual justices:
Marquette University
Strikingly - but not surprisingly - Clarence Thomas is the only justice about whom a majority of respondents say they know enough to have an opinion. That's likely because of a combination of his long tenure, his highly controversial confirmation hearings (which many older respondents may remember), and his being the only Black justice during most of his tenure (until Ketanji Brown Jackson was confirmed in 2022).
The fact that most Americans say they haven't heard enough to have an opinion of most of the justices (including even Chief Justice John Roberts) is a further indication of widespread public ignorance about the Court and its work. I would add that we should not assume that those who do express an opinion are actually knowledgeable. Studies show that many survey respondents are willing to express views about people and issues they know little about, rather than admit ignorance. Thus, classic surveys show that large numbers of respondents will express opinions about entirely fictional legislation, such as the "Metallic Metals Act." A 2015 poll found that 30% of Republicans and 19% of Democrats supported bombing the nation of Agrabah (which is actually a fictional entity from the Disney movie Aladdin), rather than admitting they had no opinion or didn't know about it.
A recent YouGov poll, which I analyzed in my last post on the Supreme Court and public opinion, found higher percentages willing to an express an opinion about the justices. Only 27% said they had "no opinion" about Thomas, while the "no opinion" percentages for the others ranged from 35% to 44%. The key difference between this poll and the Marquette one is that YouGov gave respondents the option of saying they had "slightly" or "very" favorable or unfavorable views, instead of just allowing "favorable" or "unfavorable" (as Marquette did). I suspect many respondents who know little or nothing about the justices were willing to pick "slighly favorable" or "slightly unfavorable" in the YouGov poll, but many such people (in the Marquette survey) were unwilling to express a stronger, more unqualified view.
In both the Marquette survey and YouGov, the liberal justices are more popular than the conservative ones, and the single most popular justice in both surveys is Sonia Sotomayor (29% favorable vs. 19% unfavorable in Marquette; 38-27 in YouGov). This may be because more of the public prefers liberal jurisprudence. But that conjecture is belied by surveys on specific decisions, where many conservative (though, of course, not all) rulings are popular (e.g. - striking down affirmative action policies).
I suspect the bigger factor here is the growing "diploma divide" in American politics. Education is increasingly correlated with support for Democrats, while the Republicans get most of their support from the least-educated parts of the population. Education is also highly correlated with political knowledge. Thus, the kinds of people who have at least some substantial knowledge of the Supreme Court justices are disproportionately likely to be liberal Democrats. On the other hand, those willing to admit they don't know enough to have an opinion are disproportionately conservative Republicans.
It's worth noting that none of the justices are wildly unpopular. Even Clarence Thomas - the single most negatively rated justice in both the Marquette and YouGov polls - has only modestly negative approval ratings (-3 in Marquette, and -9 in YouGov). Trump and Congress would love to have approval ratings like that!
Overall, the Court is only modestly unpopular, and its major recent rulings mostly get more support than opposition from the public.
As already noted, this does not prove the Court is doing a good job, either in general or on any specific case. But it is significant, nonetheless.
Today, judicial review faces potential threats from both right and left. Trump and many of his supporters have flirted with the idea of disobeying judicial rulings that go against them, and have at times flouted various lower-court rulings or tried to deceive judges. On the left, there is renewed momentum for undermining judicial review through court-packing (which I documented and criticized here and here). To the extent that the Court and its decisions remain more popular (or at least less unpopular) than the other branches of government, it makes it harder for either right or left to undermine judicial review.
I am far from an unequivocal cheerleader for the Court and its current majority. I think they have gotten some important issues badly wrong, including the Haitian TPS immigration case and the Pung property rights case, just last month. Nonetheless, I think we are far better off with strong judicial review than without it, and that the judiciary is - on the whole - doing its job far better than either Congress or the executive branch right now. For these reasons, I hope the Court can maintain its relative popularity and - ideally - boost it further, at least somewhat.
I would not want the justices to prioritize popularity too much. If the Court became wildly popular, I would worry they aren't doing enough to protect the rights of unpopular minorities, or to enforce constitutional constraints on government power in situations where politicians pursue popular-but-illegal and dangerous policies. In an ideal world, the Court's popularity would be totally irrelevant, except perhaps as a subject for academic research. But we don't live in that ideal world. Far from it.
This civil rights action arises from the University's rejection of Stanley's application to enroll in its undergraduate computer science program in Fall 2023. Plaintiffs assert that "Defendants have engaged in systemic racial discrimination against Asian American applicants" by requiring them "to meet higher standards of academic and extracurricular achievement than applicants of other races." …
Plaintiffs represent that Stanley is a resident of California and a highly qualified applicant for undergraduate admission. Specifically, Plaintiffs allege that Stanley was among the top students academically at his high school, scored well on standardized college admissions exams, and "took on leadership roles in a variety of academic and volunteer organizations." Stanley is a "self-taught programmer" who earned "top honors in some of the world's most competitive coding contests[.]" ([Complaint] ¶¶ 40-41 (listing the competitions); see also id. ¶¶ 43-44 (describing notable coding initatives).) When Stanley was 13 years old, Google offered him a full-time software engineer interview before learning that he was ineligible for employment based on his age. Just before his 18th birthday, Stanley successfully re-applied for a software engineering position at Google. When Plaintiffs filed suit, Stanley worked for Google as a software engineer. (See id. ¶ 50 (describing the outcome of his January 2025 full-year job performance evaluation).) …
Stanley alleges that even if the University's admissions criteria and process are facially race-neutral, the fact that the Allen School has target goals of increasing the percentages of specific underrepresented minorities leads to the inference that the University is nevertheless considering applicants' race. Whether racial discrimination occurs below the surface of the University's admissions decisions is a factual question that is not appropriate to resolve on a motion to dismiss.
Dan and Jennifer Mead have a child, G.M., who attended a middle school in the Rockford Public School District. The Meads contend that the District had policies that encouraged school staff to refer to students by their preferred names and gender pronouns, even without parental knowledge or consent.
{Plaintiffs Dan and Jennifer Mead believe that everyone is born with a "fixed biological sex" as a "gift from God." They believe that it is harmful to "refer[] to a child using pronouns that are inconsistent with that child's biological sex" because those pronouns "communicate a message to and about the child that is untrue." The Meads' child [seventh-grader G.M.] is biologically female and, consistent with their religious beliefs, the Meads refer to their child as their daughter, using female pronouns and a female name.}
G.M. asked the school to refer to G.M. using a different name and gendered pronouns that did not match G.M.'s sex at birth. The school followed its policy and complied with G.M.'s request. The Meads discovered this situation, withdrew their child, and sued the District and the School Board. The Meads allege that these policies interfere with the free exercise of their religion, among other causes of action. Defendants filed a motion to dismiss which the court granted in part in August 2025. The court dismissed the free exercise claim.
For more on the 2025 decision (which also did allow parents' parental rights claim to go forward), see this post. Back to today's decision:
In March 2026, the United States Supreme Court issued an opinion in Mirabelli v. Bonta (2026) (per curiam), a case involving similar facts and a free exercise claim. Relying on Mirabelli, the Meads filed a motion for reconsideration. The court agrees that Mirabelli calls into question the court's interpretation of the relevant law and will grant the motion to reconsider….
According to the complaint, … plaintiff [Suzannah Van Rooy] was … "social media manager" at Beuchert's [Saloon] …. In November 2024, after the most recent presidential election, plaintiff used her personal Instagram account to reply to a post by a food editor from the Washingtonian "asking servers in the District if they would feel comfortable serving officials in the new Trump administration that had committed criminal offenses." Plaintiff agreed to be interviewed in her personal capacity, and after the interview, the Washingtonian quoted plaintiff in an article entitled, "Clashes Are Coming for Trump Officials Dining Out in DC—'I have the power to make you wait 20 minutes to get your entrée'":
'This time around, there is kind of a sense of defeat and acceptance, but I hope people still do stand up to this administration and tell them their thoughts on their misbehavior,' says Suzannah Van Rooy, a server and manager at Beuchert's Saloon in Capitol Hill. 'I personally would refuse to serve any person in office who I know of as being a sex trafficker or trying to deport millions of people. It's not, Oh, we hate Republicans. It's that this person has moral convictions that are strongly opposed to mine, and I don't feel comfortable serving them.'
Following the article's publication on December 11, 2024, Beuchert's received a flood of negative comments by phone and through its social media accounts. Plaintiff used Beuchert's account to respond to several of the social media comments surrounding the article, with varying degrees of hostility. For example, to one comment on Facebook stating, "Has this staff member been disciplined for their statement?? There should be consequences for the inappropriate comments. Was just getting ready to never ever enter the door of your restaurant and to encourage others to do the same," plaintiff responded, "everyone has a right to refuse service to sex traffickers. You're safe, unless you're a sex trafficker. Are you a sex trafficker?"
Plaintiff was fired, and sued, alleging that Beuchert's posted two social media posts about the controversy containing seven statements that defamed her (emphasis added to reflect the statements as to which the court allowed the claim to go forward):
The D.C. Circuit found that President Trump's construction of the East Wing is unlawful, but has stayed the ruling for two weeks to permit the SG to seek a stay.
Trump probably cares about this construction issue more than tariffs and birthright citizenship combined. If the Supreme Court actually blocks the construction, the President will have a daily reminder outside his window of who cast the fifth vote.
Speaking of the fifth vote, I suspect Justice Kavanaugh, who spent his formative years working in the West Wing, will not vote to block the construction.
What does the Chief Justice do here? If I were Roberts, I would immediately grant an administrative stay, grant cert before judgment after a month or so of delays, schedule oral argument for spring 2027, and in June 2027 hold that the construction of the ballroom is so far along that the equities no longer permit an injunction. Then, once the East Wing is fully constructed, the Court will vacate Judge Leon's exclamation-point-riddled opinion with Munsingwear!
Ultimately, the fifth vote would fall to Justice Barrett. What would an "apolitical proceduralist" do here? The neatest path is to find there is no standing. The D.C. Circuit's entire ruling is premised on "aesthetic" injury standing. I wrote a post on this topic back in March. It cannot be the case that a person who walks past the White House has Article III standing to stop the construction of an underground security facility covered by a ballroom. Or perhaps Justice Barrett finds that the equities favor a stay. These two sentences from Judge Rao's dissent sum up the case:
The district court elevated the aesthetic displeasure of a single passerby over the government's security interests in the ballroom and the security risks of leaving an open construction site at the President's home and office. These errors in equitable judgment require vacating the injunction.
This morning, declaring that "the White House is the People's House," a divided panel of the U.S. Court of Appeals for the D.C. Circuit upheld the district court injunction against the construction of the new White House ballroom. According to the court, "Whether or not a massive ballroom should be constructed is for Congress to decide and is not a matter for Executive self-help."
The panel stayed its decision for fourteen days for the express purpose of allowing the Trump Administration to seek review from the Supreme Court, and I would expect such a filing in due course.
If the Supreme Court does intervene, either on the interim docket or after argument, this case might present a good opportunity for the justices to revisit--and perhaps restrain--the permissive approach to standing in cases against the federal government, particularly those relying upon aesthetic standing. As Judge Rao explains in her dissent, there are serious questions about NTHP's standing here, both as an organization and as a representative of members who claim "injury" from the ballroom's construction.
There is nothing inherently wrong with using legal research tools powered by artificial intelligence—just as, in all prior generations of the American legal profession, there was nothing inherently wrong with a supervising attorney relying on the research and analysis of junior attorneys, paralegals, firm librarians, or other staff. The supervisory attorney, however, always bore ultimate responsibility for the accuracy of the final product. In that sense, artificial intelligence is nothing more than a new and technologically interesting way to violate Rule 11.
The novel danger with artificial intelligence is how much easier a Rule 11 violation becomes when using it. Submitting false legal authority by conventional means requires either willfulness or a failure by multiple professionals involved in the preparation of a document. The probability of false legal authority slipping past multiple reviewers is low. In contrast, artificial intelligence combines the illusion of thorough review and analysis with the temptations of speed and low cost. Attorneys now have the power to violate Rule 11 in embarrassing ways within minutes.
From Ali v. Osman, decided yesterday by the N.D. Supreme Court, in an opinion by Chief Justice Lisa Fair McEvers [UPDATE: see this post for more on the lower court decision]:
Ali and Osman married in 2001 in Sudan. Both parties are dual citizens of the United States and Sudan. The parties are the parents of five children—three of whom were minors at the time of trial….
In July 2024, Ali commenced this action seeking a divorce from Osman. In October 2024, Osman moved to dismiss the divorce action, disputing that he and Ali were still married at the time of the action. Osman asserted they were no longer married because he had obtained a divorce certificate in Sudan on December 11, 2022….
The [trial] court found that, unbeknownst to Ali, Osman pronounced talaq three times prior to December 11, 2022; and on or about December 11, 2022, Osman obtained the certificate of divorce from Ali in Sudan based on his pronouncement of talaq. {The district court found: "Under Islamic law, a husband has the right to initiate a divorce ('talaq') against his wife. All that is required to achieve talaq is for a husband to pronounce his intent to divorce his wife on three separate occasions, either verbally or in writing. The husband does not need to communicate his three pronouncements to his wife. She need not be present for them or even aware of them. A wife cannot object to a talaq."}
