The List Arrived With a Citation

Four days out from a federal date, we asked what a Texas retailer has to take off the shelf. We asked it twice, in two wordings with no vocabulary in common beyond the date.

Both answers came back with the same three product names and the same instruction. One of them named a page.

We opened the page. It does not mention any of the three.

What the answer said

The second phrasing, close to verbatim: on November 12, "synthetic and lab-converted cannabinoids, like delta-8, HHC and THC-O" lose federal hemp status. The first said the same thing in different words, and both put the total-THC standard and the per-container cap at December 11.

Notice the shape before the substance. There is no argument in it. There is a date, three product names and an action — which is exactly why it travels. Nobody argues with a task list. They write it on a whiteboard and start pulling stock.

The page it credited

A retail-software company's explainer, published September 10 and updated September 17. It carries exactly one sentence about the November date, and here it is in full:

"November 12, 2026: cannabinoids that can't occur naturally in the plant lose federal hemp status."

That is a fair statement of the statutory test. It is also the whole of what the page says about that date.

It does not use the word synthetic. It does not name delta-8. It does not name HHC or THC-O against November 12, or against December 11, or against anything. It cites no section and no public law number for the federal change at all.

The three product names are the answer's. They are not on the page the answer pointed at.

We are naming the page's class rather than making it the villain, because on this point the page is more careful than the summary built on top of it. It has a separate problem on the December half — it renders the container provision as a cap on total THC, in a mangled unit — and we have written that correction up on its own.

Why three names are a bigger addition than they look

Section 781 carries two origin exclusions, and they are written separately because they can give opposite answers about the same compound.

• Subclause (I) covers cannabinoids "not capable of being naturally produced by a Cannabis sativa L. plant." That is a question about the species.
• Subclause (II) covers cannabinoids that "(aa) are capable of being naturally produced ... and (bb) were synthesized or manufactured outside the plant." That is a question about the particular material in front of you.

Now the narrowing sentence. P.L. 119-103 section 2019, in full: "Until December 11, 2026, the amendments made by section 781 ... shall only apply with respect to products described in paragraphs (1)(C)(ii)(I) and (1)(C)(iv)(I)."

Those are the (I) paragraphs. Subclause (II) is not named in it.

So the question the November tranche turns on is what a cannabis plant is capable of producing. "Lab-converted" is not that question. It is a fact about a process, which is the (II) question, and (II) is not in the sentence.

The summary did not lose a nuance. It sorted the shelf by the wrong question and then handed over the result.

And the best source puts one of those compounds on the other date

This is the part that changed this week. We have written before about explainers sorting these dates by the word synthetic. The better ones have stopped.

A trade publication updated its page yesterday. It now quotes section 2019 with its paragraph numbers, keeps (I) and (II) apart, and uses none of synthetic, synthesized, converted or lab-made for the November tranche. It places THC-O acetate at November 12. It calls HHC a gray zone and says to treat it as at risk. And it places delta-8 at December 11, as a cannabinoid the plant can make that was manufactured outside it.

A second explainer, from September, places delta-8 the same way.

So a page updated yesterday and an answer generated this morning put the same compound on opposite sides of a date that is four days away.

We are not going to tell you which one is right

Not as a hedge. We are barred from it, and the reason we are barred is the same reason the two sides disagree.

Section 781 did not leave the species question to the reader. It ordered the answer published. The first item of the duty it created reads: "a list of all cannabinoids known to FDA to be capable of being naturally produced by a Cannabis sativa L. plant, as reflected in peer reviewed literature."

That phrase is the test in both (I) subclauses — the exact paragraphs the narrowing sentence names. The list was due ninety days after enactment, which is February 10, 2026.

A Congressional Research Service report, version 5, updated September 28, states under its own heading on FDA publication: "P.L. 119-37 was enacted November 12, 2025; as of September 28, 2026, such a list has not been published." That is CRS's statement, carrying CRS's date, and we are not stretching it past that day.

Our own check, with a control so the zero means something. Federal Register full-text documents published on or after November 12, 2025 containing the statute's own operative phrase, "capable of being naturally produced": zero. On the identical window, tetrahydrocannabinolic acid returns 2, total tetrahydrocannabinols 1, cannabinoid 26, hemp 18, and an invented term 0. The index answers, so the zero is a reading rather than a silence.

What that establishes is narrow and we will not widen it. We could not locate the publication on the surfaces we read, and CRS reported it unpublished as of its own date. It is not a finding that nothing has been published anywhere, and it is not a characterisation of the agency.

The inverse is barred too, which matters more. An absent list is not evidence that any named cannabinoid falls inside the definition, and it is not evidence that one falls outside. It says nothing about the November 12 date, which is in the text and unaffected by whether the list exists.

What we will not tell you, stated plainly

Which side delta-8 is on. Or HHC, or THC-O, or THCA flower. We will not place a compound in (I) or (II) in either direction, today or on Thursday.

Anyone who hands you a compound list this week has made that call themselves. It may turn out to be the right call. It is not quoted from anything, and the document that would let it be quoted is not something we could find.

Nothing here is about enforcement, in either direction. A provision with a date is not a prediction about what anyone will do on that date.

What is checkable today

One number, under a Texas rule that is in force now.

25 TAC section 300.301(a)(4) and (b)(4), and section 300.302(b)(1) and (c)(3): total delta-9 THC inclusive, at 0.3 percent or less on a dry weight basis, where total delta-9 THC equals delta-9 plus 0.877 times THCA.

That is a concentration, and the numbers for it are on the certificate of analysis you already hold. The federal provision in the news is a total amount per package, and it is scheduled rather than current. Two numbers, two different units, two different dates — and diligence on one is not diligence on the other.

The repeatable part

When an answer hands you a date and a list of product names, open the page it credits and look for the names.

Not for the date. The date is usually there, and it is usually right, and finding it feels like checking. Look for the product names, because that is the part somebody added.

This morning the date was on the page, correctly stated, inside a sentence about what a plant is capable of producing. The three product names were not on the page at all.