In the span of a few weeks, the federal government finalized its clinical trial guidance for psychedelic drugs, scheduled a public hearing on their future therapeutic use, signed an interagency agreement between HHS and the VA to prepare veterans’ care for products that do not yet exist, opened funding competitions for ibogaine research, and watched the Massachusetts House pass its own state-level pilot program. Read together, it looks like a field crossing a threshold. It is not. Nothing in this run of announcements moved a single one of the actual scientific questions this desk has spent the year tracking, and treating institutional momentum as evidence of scientific momentum is exactly the mistake that has burned this field before.
What actually happened, stated plainly
Every action in this stretch is a preparedness or process action, not a validation action. Final guidance tells sponsors what the agency expects a trial to look like; it does not tell anyone whether a given drug works. A public hearing collects input; it does not deliver a verdict. An HHS-VA memorandum of understanding builds workforce and data infrastructure contingent, explicitly and repeatedly, on approvals that have not happened. A House-passed state pilot program would permit three clinics to participate in a five-year research exercise, assuming it clears the Senate and the Governor; it is not law yet, and it is not a medical endorsement. None of these documents contains a finding. They are the government organizing itself to act quickly once a finding exists, which is a different thing entirely from a finding existing.
The specific questions that did not move
This desk has spent this year identifying the same handful of unresolved problems across nearly every major psychedelic trial it has covered, and none of them were touched by anything in this announcement run.
Functional unblinding is still functional unblinding. The FDA’s own final guidance spends real, specific attention on blinding questionnaires, expectancy evaluations, and active-comparator designs, precisely because the problem remains unsolved. A regulator writing detailed instructions for managing a confound is not evidence the confound has been managed. It is evidence the agency has finally admitted, in writing, how serious the problem still is.
The durability question sitting inside Compass’s own COMP006 data is untouched. A majority of the 25 milligram arm received retreatment after Week 9, and the six-month durability claim rested on a post hoc, observed-data analysis, facts this desk verified directly against the company’s own investor materials. Nothing in the guidance, the hearing announcement, or the MOU changes what that data shows. The new guidance’s insistence on prespecified retreatment criteria and blinded long-term follow-up is, again, a regulator responding to exactly this ambiguity, not evidence it has been resolved in any specific drug’s favor.
The 5-HT2B cardiac question remains open. The guidance now requires specific microscopic valve evaluation when a candidate shows 5-HT2B agonist activity, precisely the mechanism this desk has flagged as an unresolved risk for the field’s non-hallucinogenic and high-potency candidates. A required test is not a clean result.
The precedent this field should already have learned from
This is not a hypothetical risk. It already happened, in this same story, to this same field, and the evidence sits inside this desk’s own reporting. Resilient Pharmaceuticals, the company formerly known as Lykos, had every political advantage heading into the FDA’s first round of priority vouchers this spring: a sympathetic administration, a veteran-PTSD indication the White House had explicitly prioritized, and the longest clinical track record of any psychedelic developer. The agency passed it over for a less mature program with a cleaner methodological profile. Political and institutional momentum did not purchase regulatory forgiveness for functional unblinding, safety-reporting concerns, and durability gaps the agency had documented for years. The FDA’s own final guidance and hearing announcement are, if anything, a doubling down on exactly the standards that sank that program, not a signal that the standards are loosening for whoever moves next.
What this actually is, and why it still matters
None of this is an argument that this stretch of federal activity is empty theater, or that it should be dismissed. A finalized guidance document that sponsors can actually build against, after three years of designing around a draft, is genuinely useful, and this desk has covered its specific value in detail. Interagency infrastructure that trains clinicians and builds real-world data systems before an approval, rather than scrambling after one, is a defensible use of the years a review process actually takes. The federal government organizing itself around a drug class it has not approved is a legitimate story, worth the attention this desk has given it.
The argument is narrower and more specific: readiness to deploy a drug quickly is not evidence the drug is more likely to be approved, and conflating the two produces bad judgment in exactly the places judgment matters most, capital allocation, trial design decisions, and how patients and advocates calibrate their expectations. A sponsor reading this month’s news as license to relax on blinding rigor or durability evidence is reading it backward. A reader treating five pieces of process news as five reasons psilocybin or MDMA therapy is closer to their medicine cabinet is making the same error a market made this spring, when a Polymarket contract on federal psychedelic approval swung from 18 percent to 68 percent on the April executive order alone, a political directive that arrived days before any concrete regulatory action, agency review clock, or voucher award, and changed nothing about the underlying evidence.
What we think readers, sponsors, and investors should actually watch
The leading indicator was never going to be the pace of federal announcements, and this stretch of news does not change that. It is, and remains, whether a specific trial’s blinding was addressed by design rather than gestured at, whether a durability claim rests on a prespecified retreatment framework rather than a post hoc analysis built after the fact, and whether a cardiac safety question has an actual answer rather than a newly required test. Those are boring, specific, trial-by-trial questions. They are also the only questions that have ever actually decided whether a psychedelic drug gets approved, and nothing about the volume or speed of this month’s federal activity changes that. The government moving fast is real news. It is not the same news as the science moving at all, and a field that keeps confusing the two is going to keep being surprised by the same outcome.