The United States is edging closer to a decision on the case for rescheduling cannabis. Yet the strongest evidence against that move may be coming from an unexpected source: the federal government’s own paperwork.

Chief Administrative Law Judge Derek Julius is preparing his recommendation in the marijuana rescheduling proceeding. Dr Heidi Overton, meanwhile, awaits Senate confirmation as the next Commissioner of the US Food and Drug Administration. Buried inside the Drug Enforcement Administration’s 2025 National Drug Threat Assessment sits marijuana rescheduling evidence that regulators, and the public, cannot easily set aside.

What the DEA Actually Found

The DEA’s own assessment found that THC potency keeps climbing, and that potency remains largely unregulated in states that have legalised marijuana. More troubling still, the agency stated plainly that product labelling “is not controlled and is often significantly inaccurate”. That puts every consumer at risk, and children face an even greater one, since many could mistake THC edibles for ordinary sweets.

This is not a fringe claim; it sits inside a federal threat assessment, and it forms a central plank of the marijuana rescheduling evidence now working its way through the courts. The same report describes illicit cultivation sites using banned pesticides, licences obtained through falsified means, and production limits routinely exceeded by operators who launder proceeds through straw owners.

To be fair, most of the DEA’s findings concern illicit trafficking rather than every product sold legally. But the report still establishes something in the government’s own words. Legalisation has not solved the problems of uncontrolled potency, misleading packaging or inaccurate labelling.

A System That Keeps Failing Its Own Checks

Massachusetts shows why nobody can dismiss this as purely a black market problem. In 2025, the state’s Cannabis Control Commission issued public health advisories. Marijuana flower and pre rolls had already passed required compliance testing. Inspectors later found them contaminated, or potentially contaminated, with yeast, mould and coliforms above legal limits.

These were not products bought from an unlicensed dealer. They moved through licensed retailers and medical treatment centres, tracked from seed to sale. Only afterwards did the Commission suspend the licence of the laboratory responsible for testing them. Its August 2025 advisory named products connected to 544 lab samples that had already failed yeast and mould testing once.

“These were not products purchased from a street dealer,” said Duane Boise, Chief Executive Officer of MMJ International Holdings, whose company has spent years pursuing pharmaceutical grade cannabinoid development through federal channels. “They moved through licensed businesses, state required testing and tracked inventory. The system still failed to identify the problem before products reached consumers and medical patients.”

Emergency Rooms Are Left to Fill the Gap

The consequences show up clearly in hospital data. Connecticut’s Poison Control Centre has logged 668 cannabis related calls involving minors since 2019. Of those, 280 involved children under six. Michigan treated more than 400 children under six in emergency departments for marijuana ingestion between 2020 and 2022. Around 80 of them needed critical care. North Carolina recorded 3,542 cannabis related emergency department visits in 2025 alone. More than a third involved people aged 24 or younger.

In October 2024, a Wisconsin restaurant shared a kitchen with a hemp derived THC edible manufacturer. Staff accidentally served THC infused oil to unsuspecting diners. The Centers for Disease Control and Prevention identified at least 85 people with symptoms of THC intoxication. Twenty eight of them needed emergency treatment.

“The emergency department has become the quality control function for a marketplace that often cannot tell a physician what the patient consumed or in what quantity,” Boise said.

Passing a State Test Is Not the Same as Proving a Medicine Works

Here lies the heart of the matter, and the reason it deserves closer scrutiny. State cannabis testing typically checks a limited panel, covering potency, selected pesticides, microorganisms, solvents and heavy metals. That panel does not establish complete chemical characterisation, nor does it confirm consistency between manufacturing batches, validated production processes or safe dosing. Drug interaction risks and evidence from properly controlled clinical trials sit entirely outside its scope.

For any medicine regulated by the FDA, quality has to be built into the manufacturing process itself, rather than confirmed afterwards by testing a single sample. That distinction, between testing a product and actually controlling it, is where much of the marijuana rescheduling evidence becomes difficult to argue around.

What Federal Scientists Told the Hearing

Parties filed post hearing briefs in the DEA’s rescheduling proceeding on 17 August 2026. Those briefs describe testimony from FDA official Dr Dominic Chiapperino. He reportedly called marijuana “a variable substance” without “known and reproducible” chemistry. That means it would not meet the first element of the DEA’s traditional five part test for accepted medical use. DEA pharmacologist Dr Luli Akinfiresoye reportedly went further. She testified that marijuana failed to satisfy any of the five elements, from adequate safety studies through to acceptance by qualified experts.

If chemistry genuinely is not reproducible, doctors face a real problem. They cannot know whether one batch of a product contains the same constituents, in the same concentrations, as the last. That is not a technicality. It is the difference between a medicine and a guess.

A Legal Ruling Cannot Rewrite the Chemistry

Judge Julius will eventually issue his recommendation. Whatever framework the Department of Justice’s Office of Legal Counsel favours, no scheduling decision can remove microbial contamination, correct an inaccurate label or supply the missing toxicology and clinical trial data. A recommendation is not a prescription, and a state licence is not FDA approval.

MMJ International Holdings points to its own experience as a contrast. The company has spent more than eight years and over ten million dollars developing standardised cannabinoid capsules through federal pathways. That work includes two investigational new drug applications, for Huntington’s disease and multiple sclerosis. Its products remain investigational, unapproved and under a full clinical hold. That is precisely because the FDA has required proof of consistent manufacturing and dosing before anyone receives a single capsule. It is worth noting that MMJ is also a petitioner in litigation over the rescheduling order itself, so its call for tighter standards is not without commercial interest.

“We are asking why the word medical is permitted to do the work that chemistry, manufacturing controls, toxicology and clinical trials are supposed to do,” Boise said. “Calling cannabis medicine does not make it medicine.”

What the Marijuana Rescheduling Evidence Means for the Next FDA Commissioner

Dr Overton’s confirmation outcome remains uncertain. But the marijuana rescheduling evidence already on the public record raises a question nobody can wave away. Should regulators treat any product as medicine before it meets the same standards asked of every other medicine? That question matters most for the children who remain among those most exposed to the risks, and it is why the case for rescheduling cannabis deserves more scrutiny, not less, before the ink dries on a final decision.

Source: DEA Has Published the Evidence Against Marijuana Rescheduling