They’re So Back: A few predictions for the first year of the second Trump government

Annual predictions are weak and boring. And I never claimed to be original, so here are mine. Really, like everyone else who does these, I’m looking for an outlet to catalogue the things I’ll be paying close attention to in the new year, a yardstick to measure against at the end of the year, and an opportunity to brag a lot if the more out-there takes come true.

The TikTok ban is unconstitutional: When Congress passed the Protecting Americans from Foreign Adversary Controlled Applications Act,^[1] I felt that it was obviously unconstitutional—especially the provisions that apply solely and expressly to TikTok—and I assumed mine would be the mainstream take. That seems not to be the case.^[2] So maybe it’s a hot take but the DC Circuit’s opinion does not hold up, and government’s position in its opening brief, even less so. Ultimately, the government’s (‘Our anti-propaganda law is not contend-based so it’s fine.’) doesn’t pass the smell test, and the Supreme Court is not going to endorse the DC Circuit’s bootlicking, blind adherence to Congressional and Executive Branch judgment.

The tech industry splits with Trump: I know I’m not breaking new ground with this one, and we’re already starting to see it in the dust-up over conservative attacks on the H1B system. But it was obvious even before the election that Trump’s victory required him to build an untenable coalition made up of virulent, populist racists and mega-rich globalists. It’s not just Trump, though. The collapse of college-educated support for the Republican Party put it on this path a long time ago. White, non-college educated voters are not sufficient in number to support Republican majorities, and despite some inroads with Latinos and (minimally) with young black men, those gains are not enough in the medium-term either. So Republicans has relied on the mega-rich and their worst lackeys—car dealership owners, metro-area Chick-Fil-A franchisees, etc.—to replace their lost college-educated men. But the interests of the latter group are not reconcilable with those of the working, Appalachian whites that are the party’s strongest soldiers. That irreconcilability finally started making the front page in the 2024 election, and I expect it will start to crack in the coming year.

Continued weakening of qualified immunity, but no action from Congress: I have no data to prove this, but I follow the federal courts of appeals more closely than most people, and I have the distinct sense that circuit judges have taken a more critical eye to grants of qualified immunity since 2020. I blame this on two things: (1) the doctrine became much more salient in the wake of the 2020 uprising following the death of George Floyd, and (2) the Supreme Court has indicated that it looks on the doctrine somewhat less favorably that it has in the recent past.^[3] I think we will see another round of this in the coming year. As with the reaction to the death of Georgia Floyd, liberal and left-wing outrage will be more vigorous and faster than under normal administrations. And somewhere in America, police officers will unjustly murder someone (because they do that every year). So, once again, qualified immunity will rise in salience, and courts will be driven to take an even more searching look at police officers’ actions when they get the chance.

The AI bubble pops: This might seem like an out-there take, but the stars are aligning for the AI bubble to reach an early and dramatic pop. 2025 is year OpenAI is converting into a for-profit company. And if the past is prologue, it will cap the year with about 5 billion in losses. At the same time, on day I wrote this post, Bluesky was blowing up over the comedic horror that is Meta’s new AI Instagram accounts. Generative AI products are finally making their way into the mainstream, and it’s going to be a disaster.^[4]

The Supreme Court will grant cert. on whether it should reconsider McDonnell Doulgas:^[5] This one is kind of a stretch, but in the courts of appeals, the McDonnell Douglas approach to analyzing employment discrimination claims is already nearly a dead letter. I mean, it comes up in every case, but its dispositive in any. That’s for a few reasons: (1) the courts have already explicitly limited its scope to one very narrow context—at summary judgment;^[6] (2) the appellate courts have increasingly relied on non-McDonnell Douglas analyses to find triable employment claims.^[7] As Judge Newsom convincingly (if somewhat self-satisfiedly) explains, the framework is made up, contrary to the text of Title VII and Rule 56, and increasingly useless.^[8] And the framework already has at least two enemies on the Court.^[9] So it’s ripe for overturning. That said, given that the Court takes ever fewer cases, I will certainly be bragging if this one comes true.

Elsewhere on the internet

Meshal v. Comm’r, Ga. Dep’t of Pub. Safety, No.23-10128 (11th Cir. Sept. 16, 2024): A case in which the Eleventh Circuit affirmed the denial of qualified immunity to Georgia police officers who detained a trucker solely on the basis of his presence on No Fly List. The conclusion that extending a traffic stop or searching a person’s truck solely on the basis of his presence on the No Fly List violates the Fourth Amendment is not all that interesting. Being on the No Fly List is, obviously, not a crime.^[10] But the determination that this was clearly established is worth reading, if only because it cuts against the Eleventh Circuit’s normal assumption that police are dog shit stupid.

Stalley v. Lake CI Warden, No. 22-10881 (11th Cir. Dec. 30, 2024): A reality check on one of my predications above: in this case, the Eleventh Circuit affirmed the grant of qualified immunity to various prison officials involved in the death by asphyxiation of an inmate at Lake Correctional Institution in Florida, who was high on synthetic cannabinoids and tried to fight the guards offering him medical aid. The main opinion is notable for complete dishonesty, claiming to apply Rule 56’s plaintiff-friendly standard, and then taking every possible inference in favor of the correctional officials. What’s perhaps more interesting, and horrifying, is Judge Carnes’s concurrence, in which he argues that no decisions prior to the Eleventh Circuit’s Wade v. McDade^[11] en banc decision (which I wrote about here) can satisfy the ‘clearly established law’ prong of the qualified immunity analysis. Meaning there is essentially no condition of confinement that could result in liability for prison officials and prisons.

Mia Sato, The influencer lawsuit that could change the industry, The Verge (Nov. 26, 2024): Worth reading if just for the drama of it all. But the idea that one influencer who hawks Amazon garbage could bring a copyright claim against another for copying her beige existence is really one of the funnier uses of copyright law I’ve ever heard of. There are landlord everywhere, for those with eyes to see.

Ohio Telecom Assoc. v. FCC, No. 24-7000 (6th Cir. Jan. 2, 2025): I didn’t put this as a prediction because it too obvious, but in the wake of Loper Bright,^[12] conservative courts of appeals—particularly the Fifth, Sixth, and Eleventh Circuits—are suddenly going to discovery that every regulatory statute passed by congress was either (1) entirely deregulatory or (2) aimed solely at regulations that protect monopolists to the detriment of workers, new entrants, and the public health. In this case, the Sixth Circuit decides that the FCC cannot regulate internet service providers, because they provide “information services” rather than “telecommunications services.” I will not try to explain the reasoning here (and if you can make it make sense, please email me).