The Second Amendment guarantees “the right of the people to keep and bear Arms,” but the right has never been absolute. The federal Gun Control Act of 1968 bars several categories of people from possessing firearms, and most of those bars draw little objection – few quarrels with disarming convicted felons. On June 18, 2026, however, the U.S. Supreme Court held that one of those restrictions reached too far: the prosecution of a marijuana user under the statute’s “unlawful users” provision could not be squared with the Second Amendment.
The decision in United States v. Hemani is worth attention not because it is sweeping – it is deliberately narrow – but because of how the Court reasoned its way there. The opinion is a working demonstration of what the Bruen historical-analogue test now demands of the government, and it should give Congress pause before it assumes that any “dangerous person” rationale will carry the day.
The Case
The matter began in Texas. Ali Hemani, a dual citizen of the U.S. and Pakistan, came under FBI investigation on suspicion of ties to terrorism. Agents executed a search warrant at his family’s home in 2022 and found a Glock 9mm handgun, roughly 60 grams of marijuana, and 4.7 grams of cocaine. Hemani cooperated. He admitted smoking marijuana about every other day and acknowledged the cocaine was his.
The terrorism suspicions went nowhere. What the government charged instead, by grand jury indictment in 2023, was a violation of 18 U.S.C. § 922(g)(3) – the provision making it a felony for anyone “who is an unlawful user of or addicted to any controlled substance” to possess a firearm. Notably, the government did not allege that Hemani was intoxicated when he possessed the gun, that he was an addict, or that his drug use had ever made him dangerous. It relied on his status as a regular user, full stop.
The district court dismissed the indictment as an unconstitutional application of the statute, and the Fifth Circuit affirmed. The government took the case up.
A Unanimous Judgment, but Not a Unanimous Court
It is tempting to call the decision unanimous, the Court affirmed 9-0, but the agreement is thinner than the vote count suggests, and the distinction matters for anyone trying to predict where the law goes next. Justice Neil Gorsuch wrote for a seven-member majority. Justice Clarence Thomas concurred separately, as did Justice Ketanji Brown Jackson (joined by Justice Sonia Sotomayor). Justice Samuel Alito concurred only in the judgment, joined by Justice Elena Kagan – meaning two Justices agreed Hemani should win but declined to sign on to the majority’s reasoning. A unanimous outcome resting on a fractured rationale is a weaker precedent than a unified one, and the splintering here is a signal that the doctrinal questions remain alive.
Equally important: the holding is as-applied, not facial. The Court did not strike down § 922(g)(3). It held that this prosecution of this defendant could not stand. The statute remains on the books.
The Framework
The majority worked within the now-familiar Bruen structure. The Second Amendment’s text presumptively protects conduct it covers; to rebut that presumption, the government must show its regulation is consistent with the nation’s historical tradition of firearm regulation. The government need not produce a “historical twin,” only a regulation that is “relevantly similar” to laws “well-established” in our history – consistent with the principles underlying our regulatory tradition.
The function of that test is to keep historical analogues honest. It is not enough for the government to find an old law that, viewed at a high enough level of abstraction, gestures in the same direction as the modern one. The analogue has to do real work. And that is precisely where the government’s case came apart.
The Government’s Analogy, and its Three Failures
The government anchored § 922(g)(3) in early “habitual drunkard” laws. A habitual drunkard, as early courts understood the term, was a person who for a considerable part of his time was intoxicated to the point of losing his ordinary reasoning faculties. Such laws ranged from requiring a bond for good behavior, to assignment of a guardian, to confinement in a workhouse or asylum, and even jail. The government’s theory was that these laws and § 922(g)(3) share a common project: identifying users of intoxicating substances and restricting their liberty to protect the public from unusually dangerous people.
On the surface, the comparison is plausible. The Court took it apart along three axes – who each law targeted, why, and how.
Who. The habitual drunkard laws reached people whose drinking had rendered them practically incapacitated – incapable of managing their own affairs, or a danger to themselves or others. A frequent drinker did not qualify; incapacitation did. Section 922(g)(3), by contrast, disarms anyone who regularly uses any amount of any controlled substance for anything other than its prescribed purpose. It requires no showing that the individual is ever incapacitated, much less dangerous. The difference, the Court observed, is not a minor variation, but a major difference of degree the government tried to argue away.
Why. The government justified § 922(g)(3) as protection of the public from “unusually dangerous” people likely to misuse firearms in violent crime. But the historical laws, the Court found, had little to do with categorical violence. They were rooted in concerns about vagrancy, morality, productivity, and protecting a drunkard’s family from financial ruin. That is not the purpose the government ascribes to the modern statute. An analogue that served a different end cannot validate a regulation aimed at a different problem.
How. The historical laws supplied process. A vagrant was sent to a workhouse upon conviction; assignment of a guardian or commitment to an asylum required a court proceeding. Section 922(g)(3) supplies none. It strips the right to bear arms automatically, the moment a person becomes an unlawful user and until he stops, without any pre-deprivation process. Status assigned by the government, without adjudication, does the disarming.
The Schedule Problem
The Court flagged a structural danger in tying the right to the Controlled Substances Act. Drugs move between schedules; conduct can be criminalized or decriminalized by administrative or legislative action far removed from any individualized finding of dangerousness. The point is not academic. Marijuana has been moving – rescheduled at the federal level, legalized in many states, with federal prosecution curtailed. A regime in which the government can designate a group as dangerous, and thereby disqualify its members from gun ownership, by adjusting a drug schedule could hand the government a power that, in the majority’s words, would risk allowing it to “quickly swallow” the Second Amendment.
That concern is the through-line of the opinion. The objection is less to disarming dangerous people than to letting the government conjure the “dangerous” category by fiat, without proof and without process.
What the Court Did Not Decide
The majority was unusually explicit about the limits of its holding, and those limits deserve emphasis because they are where the real fights will happen. The Court did not address banning addicts, or those presently intoxicated, from possessing firearms. It did not address prophylactic laws Congress might enact after finding that users of a particular drug pose a special risk. It did not touch § 922(g)(1)’s disarmament of convicted felons. And, critically, it did not decide whether the government could prosecute under § 922(g)(3) with individualized proof that a defendant’s drug use makes him a danger, or proof that a given drug always renders its users dangerous.
What the Court held is narrow and specific: the government cannot make someone a felon for keeping a firearm while using marijuana a few times a week on his status as a user alone.
What it Means
In my view, the reasoning extends naturally to recreational users generally, even though the quantity found in Hemani’s home arguably exceeded a recreational amount – the majority’s logic turns on the absence of any individualized showing of dangerousness, not on quantity. The government’s path forward, if it wants one, runs through individualized proof or a properly supported, drug-specific legislative finding. The blanket status-based ban, applied to a casual user with no record of danger, is what failed.
The larger lesson is methodological. Hemani shows the Bruen test functioning as a genuine constraint rather than a rubber stamp. An analogue must match on who, why, and how; a loose thematic resemblance will not do. Legislatures that want to regulate firearm possession at the margins now have a clearer – and more demanding – template to satisfy. That is a meaningful development, and it arrived, fittingly enough, out of Texas.
This article was written by W. Franklin Milam, an attorney in the Nashville office of Spencer Fane. For more information, visit spencerfane.com.
Click here to subscribe to Spencer Fane communications to ensure you receive timely updates like this directly in your inbox.
