UK ASA and CAP Code rules for social media.

If you run a UK business, work with influencers or manage social media for brands, the advertising rules decide what you can post, how you must label commercial content and what happens when you get it wrong. This guide walks through all of it: the core principles of the CAP Code, when a post counts as an advert, the labels that work and the ones that do not, each platform's own tools, and what has changed since the consumer law parts of the Digital Markets, Competition and Consumers Act 2024 (the DMCC Act) came into force in April 2025.

We have checked every rule in this guide against the current guidance from the ASA, CAP, the CMA and GOV.UK, and linked the source at the end. Where the guidance changed since we first wrote this article in May 2026, the text now says so. This is general guidance for marketers, not legal advice; for a specific campaign with real money or risk attached, speak to a qualified advertising lawyer or use CAP's own advice service.

What the ASA and the CAP Code are

The Advertising Standards Authority (ASA) is the UK's independent advertising regulator. It enforces the UK Code of Non-broadcast Advertising and Direct and Promotional Marketing, known as the CAP Code, which sets the rules for advertising outside television and radio. The Committee of Advertising Practice (CAP), made up of the advertising industry itself, writes the code. CAP was formed in 1961 and published the first code that year; it set up the ASA as the independent regulator in 1962.

The ASA cannot fine anyone, and it is not a court. What it can do is investigate complaints and its own monitoring, publish rulings that name the advertiser, have adverts changed or removed, and refer the persistent cases to bodies that do have legal powers: Trading Standards and, increasingly, the Competition and Markets Authority (CMA). It is that second layer, the law sitting behind the code, that changed most in the last two years.

Social media has been inside the ASA's remit since 2011. CAP announced in 2010 that the code would extend to advertisers' own websites and to other space they control online, such as their social media pages. That is the reason your brand's own Instagram grid and LinkedIn page are covered, not only the posts you pay to promote.

The code rests on four words you will have heard before: advertising must be legal, decent, honest and truthful. Those four words cover everything from a misleading health claim and an unsupported comparison with a competitor to the disclosure rules for influencer content and paid social.

What the CAP Code covers on social media

The code applies to marketing communications: content that is directly connected with selling goods, services, opportunities or gifts. On social media that takes in a brand's own posts promoting what it sells, influencer posts made under a commercial arrangement, affiliate content, content on the brand's own website and channels, and paid ads in every format.

It also takes in content that started life as somebody else's post. Genuinely organic content, where a customer posts about you entirely on their own initiative and you do nothing with it, sits outside the code. The moment you repost it, share it, put it in a highlight or use it in an ad, you have adopted it as your own marketing, and the claims in it become yours to stand behind.

A practical example. A customer tags your brand in an Instagram Story saying your moisturiser "cleared my skin in a week". If you leave it alone, you have no obligation. If you reshare it to your own Stories, you are now making that claim to your followers, and you need evidence that the product can do it. Without that evidence, the safer move is to thank the customer in a reply and not reshare.

What the code does not cover is just as useful to know. Editorial content you have no commercial link to, news, opinion and genuinely independent reviews sit outside it. So does content about matters of public policy that is not selling anything. The test is always the same: is the content selling, and is a marketer behind it?

The four principles in practice

The four words are easy to remember and easy to underestimate. Here is how each one shows up on a feed.

Claims you can stand behind

Every factual claim in a social ad needs substantiation: product results, ingredients, prices, comparisons and testimonials. If an influencer says a skincare product "clears skin in 7 days" and there is no evidence that it does, both the influencer and the brand are exposed. Words like "best ever", "guaranteed" and "works instantly" are read by the ASA as claims, not as harmless enthusiasm, when the context suggests they are objective.

Leaving things out can mislead as much as saying something false. Promoting a "50% off" price without saying what it is half of or when the offer ends, or presenting a paid product as a personal discovery, both fail the honesty test.

Extra care with children

Content that is aimed at children, or likely to be seen by many of them, needs more than the usual label. CAP's guidance on online marketing to children under 12 says that where an ad is immersive or woven closely into the content around it, the disclosure should be prominent, interruptive and make the commercial intent clear. A small "#ad" in a caption that a seven year old will never read does not do that job. If your audience skews young, say it out loud on screen as well as in text.

What triggers the disclosure requirement

This is where most brands and influencers go wrong. The duty to label content as advertising is not limited to cash. Between the CAP Code and consumer protection law, any of the following can create it:

1. Payment of any kindA fee, a commission, a revenue share or a bonus tied to results.
2. Gifted productsWhether or not the brand asked for a post. The GOV.UK guidance for creators says the duty applies to gifts "even if there's no obligation" to post.
3. Loaned productsA car for a weekend, a dress for an event, a gadget to review and send back.
4. Free services and experiencesHotel stays, restaurant meals, travel, treatments, tickets and event access.
5. Affiliate links and discount codesAnything that earns the creator money or another benefit when followers buy.
6. Discounts, leases and other rewardsThe government's guidance lists "money, commissions, discounts, leases or loans free of charge" as incentives, alongside gifts.
7. Your own businessA creator promoting their own products, services or brand is advertising, and should make that clear unless it is obvious from the account itself.

Two layers of rules are at work here, and it helps to keep them apart. Consumer protection law, now in the DMCC Act and enforced by the CMA, requires disclosure whenever there is a commercial relationship that the audience would want to know about. The CAP Code applies in full when there is both payment, in money or in kind, and some control by the brand over the content. Control is read broadly: a brief, key messages, approval before posting, a required hashtag or a right to ask for changes all count.

A gift sent with no strings attached usually falls under consumer law only. That still means the creator must make the relationship clear if they post about it. A paid post with a brief falls under both, and is an advert for every purpose.

Labels that work, and labels that do not

The point of a label is that the audience knows it is looking at an advert before it engages, not after reading to the end. The ASA's research published in February 2026 found that 80% of people want influencers to say clearly when content is advertising, and 70% want the label visible up front. Plain labels such as "Ad" and "#ad" were among the clearest; "#gifted" and "#thanks" were among the weakest.

Two changes are worth noting if you learned these rules a few years ago. "#sponsored" now sits on the GOV.UK list of terms to avoid, updated in September 2025, so do not rely on it. And "gifted" works only as context after a proper label, never on its own.

Where the label goes

Placement matters as much as wording. The label must be upfront and prominent: at the start of a caption, before any "more" cut off; at the beginning of a video or podcast, not halfway through; on each frame of a carousel or Story sequence that carries promotional content; and never lost in a cluster of hashtags or printed in a colour that disappears against the picture. The guidance is also clear that a label in your bio does not count, and nor does a disclosure in an earlier post. Each advert carries its own.

Your own account

When a brand posts about its own products on its own account, and the account name makes the brand obvious, the commercial intent is usually clear from the context. You do not need "#ad" on every post on your company grid. The exceptions are the cases where the brand is not obvious: a founder's personal account promoting their company, a second account with a different name, or a post designed to look like independent content. Label those.

Platform by platform

Each platform now has its own disclosure tool. The ASA's 2025 report said brands, influencers and agencies are "strongly advised" to use them, and that they can be used alongside or instead of "#ad", provided the tag is clear. That softens the old advice to always use both. In practice we still add a visible "Ad" wherever the platform's tag is small or easy to miss, because the test is whether a viewer understands at a glance, not whether a box was ticked.

Instagram and Facebook

Use the "Paid partnership" label, which shows the brand's name under the creator's handle. In captions, put "Ad" before the truncation point. Stories move fast and disappear, so the label needs to sit on the first frame, in a size and colour you can read without pausing. If the content will run as a Partnership Ad, the brand's paid placement carries its own "Sponsored" marker, but the creator's original post still needs its disclosure.

TikTok

Turn on the content disclosure setting, which applies a "Paid partnership" or "Promotional content" label. The commercial intent must be clear from the first seconds of the video, before the hook has done its work. Saying "this is an ad for" at the start, or putting "Ad" on screen from the first frame, keeps you on the safe side.

YouTube

Tick the paid promotion box, which shows a notice to viewers at the start. For a dedicated video, put the label in the title or the thumbnail. For an integrated segment, say it out loud as the segment starts, not in a description viewers may never open.

LinkedIn

Our earlier version of this guide said LinkedIn had no disclosure tool. That was wrong: LinkedIn rolled out a "Brand partnership" toggle in August 2023, which puts a label at the top of the post under the author's name. Because the label can look like part of the author's headline, we also start the text with "Ad" or "Paid partnership with" and the brand's name. This matters for sponsored thought leadership, affiliate recommendations and any post where a commercial relationship exists, and LinkedIn's business audience does not change the rules.

UGC and the CAP Code

UGC sits in a nuanced place, and it matters to any brand running UGC campaigns or paying creators for content. The rules break down like this:

The last case is the one most brands get wrong in paid social. A UGC advert is designed to look like a customer's recommendation, which is exactly why it works. The ad itself carries the platform's "Sponsored" label, but the script must not claim the speaker is an ordinary customer who found the product by chance if they were paid to make the video. Our article on UGC advertising covers scripts that perform without pretending.

Reviews, testimonials and the DMCC Act

This is the biggest change for social media since we first wrote this guide. Since 6 April 2025, a set of practices are banned outright under the DMCC Act, which means they are unfair in every case, without anyone having to prove a customer was misled. Several of them land squarely on social media and review content.

1. Fake reviewsWriting, commissioning or facilitating a review that is not based on a genuine experience but is presented as if it were. Paying a creator to post a "review" of a product they have not used is caught.
2. Hidden incentivised reviewsCommissioning or submitting a review that was rewarded, with a discount, a gift or entry to a prize draw, without making that clear.
3. Misleading review informationShowing reviews in a misleading way, such as cherry picking the good ones, hiding the bad ones or inflating an overall star rating. The CMA treats ratings, counts and summaries as review information too.
4. Publishing reviews without checksA business that publishes consumer reviews must take reasonable and proportionate steps to prevent and remove fake and undisclosed incentivised ones.

The CMA has started using these powers. In March 2026 it opened investigations into five businesses over review practices, including asking staff to write reviews, offering discounts in return for five star ratings and leaving one star reviews out of ratings. For social media, the practical rules are simple: never ask staff, friends or family to post reviews as customers; if you reward a review, say so on the review; and when you share reviews on social, do not edit them or pick only the glowing ones in a way that changes the overall picture.

On our own pages, client reviews are reproduced word for word from the review sites and testimonials they came from, with their source, and we never write a client's words for them. That is the standard we would hold any brand we work with to.

Prices in social posts and drip pricing

The DMCC Act also bans drip pricing. When a post or ad quotes a price, that price must include any fees a customer cannot avoid, such as booking fees, delivery charges a customer must pay, or local taxes. Advertising "rooms from 99 pounds" and adding a compulsory resort fee at checkout is now a breach in its own right. Optional extras can still be priced separately, as long as they really are optional.

Two habits keep social posts safe here. First, quote the price a customer will actually pay, or say clearly what it does not include and why. Second, make sale claims checkable: a "was" price should be one you genuinely charged for a meaningful period, and an end date should be real. Fake countdowns and "today only" offers that run every day mislead customers just as a false price does.

Who is responsible

Responsibility is shared between the brand and the creator, and the ASA can name both. ASA rulings have held brands responsible for influencer content even when the wording of a claim came from the influencer rather than the brief.

For brands, the duty does not end when the brief goes out. Put the disclosure requirement in every influencer contract, check content before it goes live, keep records of what was sent, paid and agreed, and look at the published post to confirm the label appears as it should. If something goes up wrong, ask for it to be fixed straight away and keep a note that you did.

For creators, the duty is personal. The GOV.UK guidance makes clear that a brand's contract does not take away the creator's own responsibility to label every advert. Each piece of content carries its own disclosure; a label in the bio or on an earlier post does not cover a new one.

For agencies, the duty sits between the two. An agency that plans or places the content shares responsibility for getting it right. That is why we put the disclosure into our briefs from the start rather than correcting it after publication, and why every post on our own influencer packages is labelled as a paid partnership, with the platform tag and a visible label.

Enforcement: the ASA and the CMA

The ASA's main tools are public rulings, which name the brand and the influencer, require the ad to be changed or removed and leave a public record. In its 2025 report on influencer ads, the ASA used its AI based monitoring to look at more than 50,000 pieces of content on Instagram and TikTok. About 57% of likely ads were adequately disclosed: 55% on Instagram and 60% on TikTok. Of the ads that fell short, 80% made no attempt to disclose at all, and about one in five used a label such as "gifted", "PR trip" or "affiliate" that the ASA and CMA say does not do the job. Accounts that had been checked in 2021 had improved, from 35% adequately disclosed to 49%, but the ASA said the overall level was still below where it needs to be.

The monitoring is large scale. The ASA's annual report for 2025 says its Active Ad Monitoring system assessed 60 million ads in the year, and that 22,383 ads were amended or withdrawn. For influencers who repeatedly ignore the rules, the ASA publishes a list of names on its website, has run paid ads warning followers that the account was sanctioned, and can refer them to Trading Standards.

The bigger change is the CMA. Before April 2025 it had to go to court to enforce consumer law. Under the DMCC Act it can now decide itself that the law has been broken, and impose fines of up to 10% of a business's global turnover, or 300,000 pounds if that is higher, and up to 300,000 pounds for an individual. It can also order a business to stop, and to compensate customers. Its early focus has been on the most serious breaches and on helping businesses comply, but the ASA's own report notes that the CMA can act where a breach of the hidden advertising rules is egregious or harms consumers enough.

For a UK brand, the result is that a missing "#ad" is no longer only a reputational problem. Combined with the CMA's powers, a public ASA ruling is the kind of evidence a regulator with real fining power can use.

See content made with the rules in mind

Tap a video to play it. These are pieces our team made for brands; there is more on the work pages.

AI, green claims and less healthy food in 2026

AI generated advertising

In May 2025 the ASA and CAP set out their position on AI in advertising. There are no AI specific rules in the code, and there is no general legal duty in the UK to label AI use in ads. The existing rules apply "regardless of how content is generated, edited, or targeted". The ASA suggests two questions: would the audience be misled if the use of AI were not disclosed, and if so, would a disclosure clarify the message or contradict it?

The trap is performance. An AI generated or heavily edited image showing a cosmetic doing something the real product cannot do is likely to mislead, and adding "made with AI" does not rescue it. Routine edits that do not change what the ad says are treated much like the post-production techniques advertisers have always used. An AI avatar or a clearly comic deepfake may be acceptable when the audience can tell what it is.

Green claims

Both regulators treat greenwashing as a priority. CAP updated its environmental claims guidance in 2026 to reflect the DMCC Act, with a checklist for sign off. The principles are to think about how consumers will understand a claim, hold evidence for any objective claim, not overstate the benefit, make qualifications clear, and be open about what is a plan rather than an achievement. The ASA has named carbon neutral and net zero claims, environmental claims in food ads, and claims about disposal and recycling as priority areas. Vague words such as "eco-friendly", "green" and "sustainable" are risky on their own; absolute claims such as "zero impact" need very strong evidence. Because the CMA can now fine directly, a misleading green claim on social media carries real financial risk.

Less healthy food and drink

From 5 January 2026, ads for identifiable less healthy food and drink are banned in paid online media at any time, and on television and on demand services between 5.30am and 9pm. Many advertisers had already complied voluntarily from October 2025. The rules reach social media in three ways that matter:

CAP has suggested that a brand can engage an influencer to create content that only the brand posts on its own channels, handled with care, for example without tagging the influencer. The ASA has already issued its first rulings under the new rules, so any food or drink brand, restaurant group or hospitality business above the size threshold should check its paid social and influencer plans against the guidance before booking anything.

The compliance checklist

Run this before anything goes live. It takes two minutes per post once it becomes habit, and it covers the mistakes behind most rulings.

1. Is there a commercial relationship?Payment, a gift, a loan, a free stay, an affiliate link or a code. If yes, the content needs a label.
2. Is the label at the very start?Before any "more" cut off in a caption, and in the first seconds of a video.
3. Is it a label that works?"Ad", "#ad", "Advert", "Advertising" or a clear platform tag. Not "gifted", "collab" or "#sponsored" alone.
4. Can you see it at a glance?On a phone, against the background, without pausing or zooming.
5. Is every frame covered?Each Story frame and carousel slide that carries promotional content needs its own label.
6. Can every claim be proved?Results, ingredients, comparisons and prices, with the evidence on file before you post.
7. Is the price the full price?Any fee a customer cannot avoid is included, and sale claims are real.
8. Are the reviews genuine?No staff or friends posing as customers, rewarded reviews marked as such, and no selective editing of what you share.
9. If you are resharing UGC, would you say it yourself?Every claim in it becomes yours when you repost it.
10. Could children see it?If so, make the label prominent and say it out loud.
11. Are green claims specific and evidenced?No vague "eco" language, no absolute claims without strong proof.
12. Is food or drink in scope?If you have 250 or more staff and the product is less healthy, paid and gifted promotion online is restricted.
13. Is it in the contract and on file?Disclosure terms in writing, with records of what was sent, paid and approved.
14. Did you check it after it went live?Look at the published post, fix anything wrong at once, and keep a note.

How this fits your social media strategy

For a UK business running social media in house or through an agency, compliance is not a separate legal task that happens at the end. It is part of the content plan. Every influencer brief, gifting campaign, paid partnership, review post and piece of UGC needs the rules built in as the default, because fixing a post after a complaint costs far more than getting it right first.

That is how we work. Our social media plans include strategy, content, posting and community management, and every brief we write carries the disclosure and claims checks above. Every plan runs month to month, with 2 months notice.

Social Media Management: Seed Plan£399a month16 posts a month across 2 platforms, with a strategy behind them.

Social Media Management: Grow Plan£799a month16 posts a month, 8 of them Reels.

Social Media Management: Scale Plan£1,499a month12 Reels a month, Stories, and your comments answered Monday to Saturday.

Social Media Management: Elite Plan£2,499a monthA post every working day on up to 5 platforms, with paid social and filming days.

If you want creator content rather than management, the UGC Monthly Plan delivers three UGC videos a month from £430, with advert rights included, and our influencer packages put your brand on Velena's Instagram and Dragos's LinkedIn, every post disclosed. Before you spend anything, the free social media audit gives you a baseline of where your channels stand.

ASA and CAP Code: your questions

Dragos NistorCo-Founder and Business StrategistDragos co-founded Velena Lifestyle and leads its business strategy. Named a LinkedIn Top Entrepreneurship Voice in 2024, he writes about social media, influencer marketing and the rules around them. Velena and Dragos both guest lecture at the University of Greenwich on social media and content strategy, and are invited back several times a year.

A question this article did not answer? Ask our team, and a person will reply.