Cannabis Treaty Reform: What Real Change Could Look Like

Photo by Salya T via Unsplash.

Ask most people paying casual attention to cannabis policy what's blocking global reform and they'll point to Washington -- specifically, to the years-long slog toward federal rescheduling that finally produced a result in early 2026. That's the visible fight. But the deeper structural constraint has almost nothing to do with the DEA or Congress. It's a treaty from 1961, ratified by 186 countries, that still classifies cannabis alongside heroin and fentanyl in its most restrictive drug schedule. Every domestic legalization scheme in every country runs into that document eventually, whether legislators mention it or not.

Here's the part that isn't getting nearly enough attention: nobody has formally proposed amending the Single Convention on Narcotic Drugs. Not the US, not Canada, not Germany, not any of the countries that have legalized adult-use cannabis in open tension with the treaty's text. What's actually happening is quieter and, in some ways, more consequential -- the United States has started reinterpreting what the treaty requires, and that reinterpretation just produced a real regulatory order with legal teeth. That's the news. Not a diplomatic summit, not a vote at the UN, but a legal theory that moved from a State Department brief into an Attorney General's order in about two years.

That leaves an odd standoff at the center of international drug policy right now. The International Narcotics Control Board, the treaty's own watchdog, still says state-level and national adult-use legalization violates the Single Convention, full stop. The US government, meanwhile, just acted as though the treaty allows considerably more room to maneuver than the INCB has ever conceded. Both readings can't be fully right. So the question worth actually answering isn't whether the treaty will get rewritten -- formal amendment of an instrument this size, with this much multilateral inertia behind it, is close to a non-starter in any near-term political timeline. The real question is what reform looks like when the treaty text stays frozen but the behavior of the countries governed by it keeps moving anyway.

The 1961 Framework Nobody Wants to Reopen

Photo by Werner Pfennig via Pexels.

The Single Convention on Narcotic Drugs dates to 1961, tightened by a 1972 amending protocol, and it remains one of the most widely ratified instruments in the entire UN treaty system -- 186 states have signed on. Its core obligation is blunt: parties must limit the production, trade, possession, and use of cannabis strictly to medical and scientific purposes. To enforce that, the treaty created the International Narcotics Control Board, a quasi-judicial monitoring body that reviews national compliance and reports to the Commission on Narcotic Drugs, the UN's central drug-policy decision-making forum. The treaty also built a four-tier schedule system ranking substances by perceived harm and medical value, and cannabis and cannabis resin landed in Schedule I -- the same tier as cocaine, fentanyl, and heroin.

For nearly six decades that classification sat untouched. Then, in December 2020, the CND took its first real swing at revising it, voting 27-25 to remove cannabis from Schedule IV, the treaty's most restrictive sub-category reserved for substances considered to have no offsetting medical value. The vote followed the World Health Organization's first-ever critical scientific review of cannabis, which found enough therapeutic evidence to justify loosening that specific sub-classification. Cannabis stayed in Schedule I. The margin -- two votes -- tells you almost everything about how contested even that modest change was among member states.

What that vote didn't do is more instructive than what it did. It didn't touch the core prohibition language. It didn't reopen the treaty's text. And in the 65 years since the Single Convention entered into force, no substance has ever been formally removed from control through a full treaty amendment -- amendment procedures exist on paper, laid out in the Convention itself, but they've never actually been exercised for a substance heading toward less restriction. That's the baseline every other development in this space has to be measured against. Reform through the treaty's own front door is theoretically available and practically dormant, which is exactly why no government has bothered filing an amendment proposal. It would take years, near-universal buy-in among 186 parties with wildly different domestic politics, and a level of diplomatic capital nobody currently seems willing to spend on a fight they might well lose.

Washington's Workaround: Reading the Treaty Differently

Photo by Mikhail Nilov via Pexels.

If the front door is bolted shut, the United States appears to have found a side entrance. Since late 2023, State Department legal adviser Virginia "Patt" Prugh has been making a specific argument in front of the CND: the Single Convention's prohibition machinery was built to police cross-border trafficking, not to dictate how a country structures its own domestic drug law. That's not a call to renegotiate anything -- it's a claim about what the existing text already means. Reinterpretation, not amendment.

That argument didn't stay confined to diplomatic remarks. It reportedly informed a 2024 Office of Legal Counsel opinion inside the US government, and by April 22, 2026, it had produced something concrete: Acting Attorney General Todd Blanche signed AG Order No. 6754-2026, moving FDA-approved marijuana drug products and state-licensed medical marijuana from Schedule I to Schedule III of the Controlled Substances Act. This wasn't a symbolic gesture. The order explicitly leans on 21 U.S.C. 811(d)(1), the statutory provision that lets the Attorney General reschedule a substance specifically to keep the US in compliance with its international treaty obligations, and it cites the Single Convention directly as the legal basis for the move.

The narrowness of the order is the tell. Adult-use, recreational marijuana stayed exactly where it was, in Schedule I. If this were a blanket declaration that the treaty doesn't really bind US domestic cannabis policy, you'd expect a much broader sweep. Instead, the administration drew a careful line around medical and FDA-sanctioned products, leaving the more politically and legally fraught recreational market untouched for now. That's a reinterpretation applied surgically, not a treaty exit dressed up in legal language.

What makes this worth watching well beyond US borders is the mechanism itself. This is treaty reform executed through administrative practice rather than through the UN's amendment process -- no vote, no multilateral negotiation, no CND resolution required. It's one government's legal office deciding what its existing treaty obligations actually permit, then acting on that reading. Any other Single Convention party with a domestic legal system capable of producing an analogous opinion could, in principle, run the same play without ever going near Vienna or the UN's drug-policy machinery.

The Legal Fight Over How Far That Reading Can Go

Photo by Sasun Bughdaryan via Unsplash.

The April order didn't settle anything -- it opened a second front. On June 29, 2026, a much broader DEA administrative rescheduling hearing began before an administrative law judge, running through July 15, 2026, testing whether marijuana generally should move to Schedule III under the Controlled Substances Act, a scope well beyond the April order's narrow carve-out for medical and FDA-approved products. That hearing is where the fuller version of the rescheduling question gets litigated on the record, with expert testimony and cross-examination rather than an Attorney General's signature.

At the same time, the April order itself is being challenged in the D.C. Circuit. The core argument against it draws on NORML v. DEA, a 1977 precedent, to claim the Attorney General exceeded what 811(d)(1) actually authorizes -- that the statute lets the government reschedule to meet treaty obligations, but doesn't let it invent a split classification where medical marijuana sits in Schedule III while recreational marijuana, sold under the same drug name, stays in Schedule I. Critics call this an unauthorized hybrid schedule, a category the statute was never written to accommodate.

The international stakes of a purely domestic lawsuit are easy to miss but real. If the D.C. Circuit strikes down the reinterpretation, the US loses its own regulatory experiment and, with it, the only functioning real-world precedent other countries could point to if they wanted to try a similar legal maneuver. Reinterpretation strategies live or die partly on whether courts back them up -- a legal theory nobody's tested in litigation is a much weaker thing to copy than one that's survived judicial review. Flip it around, and a surviving order gives every other Single Convention party a tested template: narrow the treaty's reach through domestic statutory interpretation rather than touching the treaty text.

Worth flagging plainly, though: US administrative law and international treaty law are not the same body of law, and it's easy to conflate them here. A D.C. Circuit ruling resolves what 811(d)(1) permits under US statute. It says nothing binding about what the Single Convention itself requires of France, Nigeria, or Thailand. Other countries will watch the outcome closely as a data point on how far this kind of argument can travel, but nothing about the ruling itself changes their own legal obligations.

The INCB Isn't Going Along With It

Photo by Werner Pfennig via Pexels.

The treaty's own watchdog isn't buying any of it. The International Narcotics Control Board, under President Prof. Sevil Atasoy, has continued treating adult-use legalization in Uruguay, Canada, and a growing list of US states as straightforwardly non-compliant with Single Convention obligations. That position didn't soften after the April 2026 US order -- if anything, the timing makes the contrast sharper. The INCB's 2025 Annual Report launched February 26, 2026, and the 69th CND session that followed in March, both kept the board's compliance stance fully intact, just weeks after Washington had acted as though the treaty allowed considerably more flexibility than the INCB has ever granted.

It's worth being precise about what that disagreement actually means in practice, because the INCB has no enforcement mechanism that bites. It can't sanction the United States, can't impose penalties on Uruguay, can't compel Canada to reverse course. Its power is entirely reputational and diplomatic -- reports, formal recommendations, and pressure funneled through the CND. A member state that decides to act against the INCB's stated position faces disapproval, not consequences with legal force.

That gap -- a treaty body stating one thing while major member states do another on the ground -- is itself the pattern worth tracking, and it's not without precedent. Alcohol prohibition-era treaties and successive generations of international drug conventions have lost practical force well before anyone got around to formally renegotiating them; enforcement eroded first, and the paperwork caught up decades later or never did. If that historical pattern holds here, the INCB's continued objections matter less as a predictor of what will happen and more as a marker of how much resistance any eventual amendment push will face from inside the treaty system itself, not just from conservative member states at the CND table.

The Quieter Reform Track: Hemp, CBD, and Industrial-Use Exemptions

Photo by TinaKru via Pixabay.

Alongside the fight over psychoactive cannabis, a quieter and arguably more tractable reform track has been developing around hemp and CBD -- one that doesn't require reinterpreting the treaty's prohibition language at all. At a March 13, 2026 CND side event, Dr. Pavel Pachta, a former INCB board member, laid out the case for treaty flexibility that treats hemp and CBD as industrial activity governed by a different logic than narcotics control entirely.

The legal hook for this argument already exists inside the Convention. Academic work -- notably Kenzi Riboulet-Zemouli's High Compliance analysis -- points to Article 2(9), an existing exemption written into the Single Convention for industrial and horticultural uses of cannabis. The argument isn't that the treaty needs new language to accommodate low-THC hemp; it's that the exemption was there all along and has simply been under-used by governments cautious about anything cannabis-adjacent. Groups like the Transnational Institute and WOLA have pushed variations of this treaty-modernization argument for years, and none of it has reached a formal vote or adoption as an amendment -- it remains advocacy and legal scholarship, not settled policy.

What makes this track genuinely different from the rescheduling fight is the shape of the move. Rescheduling asks the world to change how it classifies a psychoactive substance -- inherently contentious, and the 2020 Schedule IV vote shows just how narrowly that kind of change squeaks through even in its mildest form. The hemp carve-out asks something smaller: define low-THC hemp and industrial cannabis products as sitting outside the definition of controlled cannabis in the first place. No reclassification, no reinterpretation of what prohibition means -- just a narrower category boundary. That's a meaningfully easier political lift, and it has textual company inside the treaty already. The Single Convention distinguishes poppy straw grown for non-narcotic industrial fiber use from opium poppy grown for narcotic purposes, showing that carving industrial applications out of a controlled-plant framework isn't a novel idea being invented for cannabis -- it's a structure the treaty's drafters already used once, for a different plant, decades ago.

What a Realistic Reform Path Looks Like From Here

Photo by Erik Mclean via Unsplash.

Run the history forward and a fairly clear shape emerges, even if the details stay fuzzy. Multilateral drug treaties have never been formally amended to loosen a schedule once a substance was placed there -- not once, in 65 years, for any drug. But they've been reinterpreted and unevenly enforced constantly. The 2020 Schedule IV vote is the proof point for both halves of that claim: it took six decades to happen, and even then it only shifted one sub-classification while leaving the core Schedule I placement untouched. That's the pace formal treaty reform actually moves at.

The more realistic near-term path is more countries quietly running the US playbook: domestic legal reinterpretation of what "exclusively medical and scientific purposes" actually requires, paired with narrow, targeted administrative actions rather than any attempt at a treaty amendment vote. Watch Canada, already fully legal and already operating in open tension with the treaty; watch Germany, which moved on partial legalization in 2024; watch Mexico, where legalization reforms have been pending in various forms for years. If any of these publish their own legal justifications built around sovereignty arguments or public-health framing -- echoing the US focus on Article 2(9) or the cross-border-trafficking reading -- that's the signal the reinterpretation strategy is spreading rather than staying a one-off American experiment.

The counter-case deserves equal weight, because it's not weak. If the D.C. Circuit litigation goes against the US reinterpretation, treaty-conservative voices at the CND -- historically Russia, China, Japan, and Singapore among the more restrictive delegations -- get a ready-made argument that the American approach was legally unsound from the start, not just politically convenient. That would slow, not accelerate, any copycat action, and it's a live possibility given the NORML v. DEA precedent the challengers are leaning on.

For businesses and policymakers trying to plan around this rather than just react to headlines, the practical takeaway is to stop waiting for a rewritten treaty. That's not coming this decade under any realistic scenario the historical pattern supports. What to expect instead is a slow accumulation of national reinterpretations -- some tested in court, some not -- that functionally hollow out Schedule I enforcement in practice while the treaty's paper text stays exactly where it's always been.

Strip away the noise and the Single Convention's actual text hasn't moved an inch, and there's no credible near-term scenario where formal amendment changes that. The 2020 vote is the clearest evidence available: two votes decided a change to one sub-schedule, after six decades of nothing, following a WHO review that took years to produce. If that's what it costs to shift a sub-classification, a full renegotiation of cannabis's Schedule I status is not a live near-term possibility, whatever momentum domestic legalization builds country by country.

The reform that's actually happening runs through a different channel entirely -- national governments deciding, on their own legal authority, what the treaty's existing language permits, then acting on that reading without asking the other 185 parties for permission. The US order of April 2026 is the first concrete regulatory test case of that approach, built on a legal argument the State Department had been making since 2023 and grounded in a specific statutory hook, 811(d)(1), rather than any claim of treaty exit.

Whether that test case becomes a template or a cautionary tale depends on litigation still working through the courts, not on diplomacy. The open question worth actually tracking isn't whether the Single Convention gets rewritten -- it won't be, not soon. It's whether the D.C. Circuit, and eventually courts or legal offices in other treaty countries, validate this reinterpretation strategy thoroughly enough that it hardens into the de facto global approach: a treaty that stays frozen on paper while national governments quietly redefine, one legal opinion at a time, how much room it actually leaves them.