The Total-THC Hemp Redefinition: What It Means for Payments

Since 2018 the federal hemp definition has rested on a single measurement: delta-9 THC at or below 0.3% by dry weight. Public Law 119-37 replaces that with a total-THC measurement, and it names THCA specifically rather than leaving the question to interpretation.

That is a narrower change than it sounds and a much larger one than it looks. Narrower, because the 0.3% concentration figure itself is not what moves. Larger, because THCA converts to delta-9 when heated, and a substantial part of the hemp-derived cannabinoid market rests on the fact that the old test measured the compound after conversion rather than the acid before it. A product can fail the new test on the same certificate of analysis that passed the old one, with no reformulation and nothing different in the bottle.

There is a second limit that draws less attention and will affect more catalogues. A hemp-derived cannabinoid product falls outside the federal definition if it carries more than 0.4 milligrams of total THC per container, measured at the innermost retail packaging rather than per serving. For finished consumer goods the container test is frequently the binding one, because a modest per-serving figure multiplied across a bottle can exceed a per-container cap while every number on the label still looks conservative.

The provision is dated 12 November 2026. On 8 August 2026 the Senate passed a continuing resolution that would move most of it to 11 December. At the time of writing that resolution has not been enacted, so the earlier date is the one to plan against rather than the one to assume.

One part is not in question. The proposed delay does not extend to synthetic cannabinoids, so delta-8, delta-10 and similar products lose federal hemp status on the original date regardless of what happens to the rest of the provision. A business whose revenue concentrates there has less runway than the headline debate suggests, and should not read the stopgap as covering it.

The honest position is that the direction is settled and the date is not. Any source stating a single date as fixed is ahead of the record. That distinction matters commercially, because a plan built on the later date fails quietly if the earlier one holds.

The most common planning error in this category is treating the effective date as the date the business needs to be ready. Payments do not work that way, and the reason is structural rather than a matter of caution.

An acquiring bank carries the compliance exposure for the merchants it boards, and card network rules oblige it to satisfy itself that a merchant’s activity is lawful. When the definition of lawful is scheduled to move, the bank’s exposure changes on the announcement, not on the commencement. Risk committees therefore reprice, re-paper, and in some cases exit a category well ahead of the statutory date, because their obligation is forward-looking even while the statute is not yet in force.

The practical consequence is that hemp and cannabinoid accounts are being reviewed now, against a definition that does not yet apply. A merchant who assumes they have until November is working to a deadline their acquirer has already moved.

Underwriting in this category has always turned on the certificate of analysis, and that has not changed. What has changed is which figure on it settles the matter.

A COA reporting delta-9 compliance answers the old question and is silent on the new one. If your certificates do not report total THC, and do not give a per-container figure for the finished product, then you cannot currently evidence which side of the new test any given SKU falls on, and neither can your acquirer. Silence is not read as compliance during a review. It is read as an unquantified risk, which is the category of finding that closes accounts.

Commissioning COAs that report total THC and a per-container figure is therefore the first action rather than the last. It is also the cheapest, because it is a testing and paperwork exercise that can be completed while the commencement date is still being argued over.

The change is not uniform, and treating it as an across-the-board threat produces worse decisions than reading it product by product.

Isolate and broad-spectrum products are generally the least exposed, because the compounds the new test captures are the ones those formats are designed to remove. Full-spectrum products need checking individually, since a total-THC reading is precisely where a full-spectrum extract differs from a broad-spectrum one. Products whose viability depends on the acid form not being counted are the most exposed, because that is the specific distinction the statute removes.

The per-container limit cuts across all three. A large-format bottle can breach a per-container cap while every per-serving figure remains modest, so container size and packaging format become compliance variables rather than purely commercial ones.

If a catalogue stops meeting the federal definition, the account does not simply continue, and it is not usually grandfathered. The sequence in practice is a review, a request for current certificates, and then either a narrowed approval covering the products that still qualify or a closure.

The part merchants underestimate is what follows a closure, because it is a cash-flow event rather than only a processing one. A reserve is held against future disputes rather than past behaviour, so it survives the account that produced it and is released only after the dispute window on the final transactions has run. A business that loses processing on the definition can find working capital held for months with no revenue running through that account to offset it.

This is why the timing of the conversation matters more than its content. A merchant who brings a product-by-product analysis to their acquirer before a review is treated as a compliance risk being managed. A merchant whose position is discovered during a review is treated as a compliance risk being discovered, and the second framing produces materially worse outcomes on reserve, on pricing, and on whether a narrowed approval is offered at all.

Processing redundancy is standard advice in high-risk categories, usually justified on the basis that banks react differently to the same merchant. That argument still holds, but a definitional change alters what redundancy can and cannot do.

Redundancy protects against a single bank’s risk appetite. It does not protect against a change in what is federally lawful, because every acquirer is reading the same statute. A second account remains worth holding, but a business planning on the basis that one of its acquirers will simply take a different view of the same definition is planning on something unlikely to arrive.

What redundancy does buy in this scenario is continuity for the part of the catalogue that still qualifies, and time to move it. That is a real benefit, and it is a different benefit from the one redundancy is normally sold on.

The useful work is not complicated, but it is ordered, and doing it out of order wastes whatever runway remains.

Test first, because no planning is possible without total-THC and per-container figures. Segment second, separating the catalogue into what passes, what fails, and what is uncertain, since the uncertain group is where the commercial decisions actually sit. Disclose third, raising it with your acquirer while it is still your disclosure rather than their finding. Then decide on the failing group, which is a business question about reformulation, packaging, channel or withdrawal, and not a payments question at all.

None of this depends on knowing the final commencement date, which is the point. Every step is worth taking whether the provision lands in November or December, and each one becomes materially harder once a review has already started.