CBD social media compliance: the rules that decide if a post is safe
A CBD post can be removed for four different reasons by four different bodies, and only one of them sends a notification. Platform policy, advertising regulators, food law and the rules of the specific market each cover part of the same caption, and a post that satisfies three of them still fails.
Updated 5 minute read
Four rulebooks, one caption
It helps to separate them, because each has a different consequence and a different speed.
| Rulebook | Covers | What happens |
|---|---|---|
| Platform policy | What may appear on Instagram: restricted goods, recommendations, commerce surfaces | Content removed, reach limited, account restricted. No warning in most cases. |
| Advertising regulation | Whether marketing is misleading, and whether commercial relationships are disclosed | Rulings, required corrections, in some markets fines. Slow, public and quotable. |
| Food and product law | Whether a hemp product may be described the way you describe it | Enforcement against the business, not the post. |
| Market specifics | Age markers, mandatory notices, wording particular to one country | Varies. Usually the first thing a local regulator checks. |
The account owner cannot pick one. A post going into three markets has to satisfy all three at once, which is why the same idea often needs three different captions.
How we split it into layers
Checking a caption against a rulebook by reading is what produces the false confidence. A rulebook is prose; a check has to be a list. We split ours into seven layers, and the ones that do not need a model are dictionaries and patterns, so the same phrase gets the same verdict twice.
| Layer | How it decides |
|---|---|
| Health and wellbeing claims | Dictionary and patterns |
| Direct selling: prices, discounts, codes, orders in messages | Dictionary and patterns |
| Hashtags: deny list, count, rotation | Dictionary and patterns |
| The rendered image: medical scenes, pills, smoking, children, before and after | Model |
| What a presenter says on camera | Model |
| Required notices for the market: age, disclaimers, AI disclosure | Dictionary and patterns |
| The brand’s own stop topics, banned words and competitors | Dictionary and patterns |
Five of those layers run without a model at all. That matters for a reason that is easy to miss: a check that gives a different answer on Tuesday cannot be tested, cannot be explained to a customer, and cannot be argued with when it is wrong.
The claim line
Most removals come from one group: wording that ties the product to a health outcome. The line is not about how careful the sentence sounds. Softening a claim does not remove it.
Do not write this
Our oil helps with anxiety and supports better sleep
Write this instead
Our full spectrum oil, 10 percent, third party tested. What is in it and how the batch was checked.
Why. Both halves name an outcome the product is not allowed to promise, and the word helps does not change that. This is the single most common phrasing on hemp accounts and the single most common reason posts come down.
The full treatment, including the phrasings that look safe and are not, is in the health claims guide.
Selling language belongs on the website
The second group is commerce behaviour on a surface that is not a shop: prices in captions, discount codes, countdowns, and taking orders in direct messages. For an ordinary retailer these are normal. On a restricted category they are the behaviour that draws review.
None of it is forbidden as commerce. It is forbidden as a post, and the fix is almost always to move the transaction to the website and leave the post describing the product.
Required notices differ by market
The third group is the opposite of a ban: something that has to be present. An age marker, a statement that the product is not a medicine, a note for people who are pregnant or taking medication, a disclosure that a visual was generated. Which ones apply depends on where the post is going.
Each market has its own page with the exact wording our engine accepts: United States, European Union, United Kingdom.
Other people's words are still your post
Reviews, customer photos and creator content carry the same rules as anything you write yourself, and in some markets they add a disclosure obligation on top. That is covered in the reviews and UGC guide.
How a check actually runs
In production the order is what saves money. Text is checked before anything is designed, because a rewrite after the artwork means the artwork is thrown away. Images are checked after they exist, because that check needs the pixels. A last pass runs before publication, when the market and the account are known.
- Before design: caption and slide text against the wording rules for the target markets.
- After design: the rendered image, for what the wording rules cannot see.
- Before publication: the whole post, in the context of the account and the market it is going to.
A phrase can be tried against the same engine on how it works. Restrictions and how they escalate are in the account restrictions guide.
Questions
- Is CBD content allowed on Instagram at all?
- Organic posts about hemp products are widely published and the category is not banned from the platform. What is restricted is paid promotion, commerce surfaces for the category in many markets, and any wording that presents the product as treatment.
- Who decides whether a caption is a claim?
- In practice several parties do, and they do not agree. The platform decides what to distribute, and regulators decide what is misleading. A caption safe from one point of view can be a problem from the other, which is why a check has to cover both.
- Does a disclaimer make a claim acceptable?
- No. A disclaimer satisfies a requirement to say something; it does not cancel a statement that should not have been made. The two are separate checks and both have to pass.
- Do these rules apply to stories as well as feed posts?
- Yes, and stories are where they are broken most often, because stories feel temporary. A story that names a condition is the same problem as a post that names a condition.
Keep reading
- Health claims on social media: describing a product without promising an outcome
What counts as a health claim in a CBD caption, why softeners like helps and supports do not help, and how to write about a product without naming an outcome.
- Why CBD accounts get restricted or removed
How restrictions escalate on hemp accounts, which behaviours trigger them, what an appeal can and cannot do, and how to keep the account recoverable.
- Reviews, testimonials and creator content on a CBD account
How to use customer photos, reviews and creator posts on a hemp account: what a repost publishes, when disclosure is required, and what to ask for instead.
- European Union: CBD rules on social media
What a hemp brand may post in the EU: novel food status, the health claims regulation, the notices every post needs and the AI disclosure requirement.