--- title: "Can you use AI to advertise my business?" url: https://canyouuseai.com/questions/advertise-my-business/ verdict: yes-with-care verdict_label: "Yes, with care" category: "Running a business" published: 2026-10-01 updated: 2026-10-01 author: "Jack Stovell" publisher: "Adapt Progress Evolve Limited" language: en-GB --- # Can you use AI to advertise my business? **Verdict: Yes, with care.** Yes, with care. AI can draft ads, captions and images, but the UK ad code applies however an ad is made, fake endorsements are out, and platforms set their own AI label rules. AI is good at the fast parts of advertising. It drafts ad copy, writes social captions and makes images. If you are a small-business owner doing your own ads on top of the stock, the invoices and everything else, that speed is tempting. The ad rules in the UK do not care how your ad was made. The CAP Code is media-neutral: if an ad is in scope, the rules apply whether a person or a tool wrote it, because ads are judged by how consumers read them. The responsibility stays with you, too. The ASA has said that even when a campaign is generated or distributed by automated methods, the advertiser keeps primary responsibility, a point it made in its Stripe & Stare Ltd ruling. "The AI wrote it" is not a defence. There is no special AI rulebook either. In June 2025 the ASA and CAP said the CAP and BCAP Codes do not contain specific rules relating to AI, so the existing rules do the work. Three traps come up again and again: images that overpromise, fake or implied endorsements, and platform labelling. ## How to do it well Start with the input. Give the AI your real facts: the product, the price, the offer terms, who it is for. Then tell it, plainly, not to add claims. These tools like to be helpful, and "helpful" often means inventing a benefit you never mentioned. ``` Write five Instagram captions for [product], priced at [price], with this offer: [offer terms], valid [dates]. It's for [audience]. Use only the facts I've listed here. Don't add any claims, and no superlatives like "best" or "number one" unless I've given you proof. ``` Fill the brackets with things that are true. Next, check every claim against evidence you actually hold before anything goes out. Section 3 of the CAP Code says marketing communications must not materially mislead, or be likely to, and the ASA says marketers must be able to substantiate their claims. If a caption says your cleaning spray removes limescale in seconds, you need evidence that backs it up. If you do not have it, cut the line. Use AI images for mood or illustration: a cosy kitchen, a sunny shop front. Do not use them to show results the product cannot deliver. The ASA says AI images used to make efficacy claims can mislead if they do not accurately reflect the product, in the same way as photoshopped or filtered images. Show the real product doing the real thing. For touching up your own product photos, see [Can you use AI to edit photos?](/questions/edit-photos/). Never generate a real person, a celebrity or a customer endorsing you. Treat it as a hard line. Read each post yourself before it goes out. The ASA notes that AI can amplify biases, so check for harmful stereotypes or anything that would not suit your audience or where the ad will appear. Finally, label realistic AI images or video where the platform asks you to. ## Rules and risks The CAP Code applies regardless of method. No AI-specific CAP rule exists, but the existing rules apply in full: misleading claims are misleading claims, and being AI-generated does not excuse harmful or offensive content. Deepfakes are the sharpest risk. The ASA says implying an endorsement that does not exist is especially problematic, and its June 2026 guidance says an AI-generated celebrity "endorsement" that looks real enough to mislead is likely to breach the Code. Personality rights and intellectual property are not regulated by the ASA, but it advises advertisers to consider the legal risk of using a public figure's likeness without permission. If you are unsure about an AI-assisted campaign, CAP's Copy Advice team can advise. On labelling, be precise. The sources here point to no general UK legal requirement to label every AI-made ad. Platforms set their own label rules. TikTok's Community Guidelines require creators to label AI-generated or significantly edited content that shows realistic-looking scenes or people, and unlabelled content may be removed, restricted or labelled by TikTok. Meta adds "AI info" labels to video, audio and image posts when it detects industry-standard AI indicators or when people disclose that they are uploading AI-generated content, and it may add a more prominent label where content risks materially deceiving the public on a matter of importance. So even if you say nothing, a label may turn up anyway. Treat AI as a fast, slightly overeager drafting assistant, not the person who signs off. You are the advertiser. This is general information, not legal advice. ## From the same studio These are our own products (the studio that runs this site makes them). - [ScriptGrain](https://scriptgrain.com/): Drafts posts and ad copy in your own measured writing style from samples you upload, so captions sound like you rather than a generic tool. - [Launch Video Maker](https://launchvideomaker.com/): Turns your business website into a short motion video with music, using only facts you approve, if you want a video ad without filming one. ## Sources - [ASA | CAP: Generative AI & Advertising: Decoding AI Regulation (31 Aug 2023)](https://www.asa.org.uk/news/generative-ai-advertising-decoding-ai-regulation.html) - [ASA | CAP: "Write a useful article helping marketers with tips for using AI" (12 Jun 2025)](https://www.asa.org.uk/news/write-a-useful-article-helping-marketers-with-tips-for-using-ai.html) - [ASA | CAP: AI and Deepfakes: Four Things Advertisers Need to Know Before They Hit "Run" (11 Jun 2026)](https://www.asa.org.uk/news/ai-and-deepfakes-four-things-advertisers-need-to-know-before-they-hit-run.html) - [ASA | CAP: CAP Code section 03, Misleading advertising](https://www.asa.org.uk/type/non_broadcast/code_section/03.html) - [TikTok: Community Guidelines, Integrity and authenticity](https://www.tiktok.com/community-guidelines/en/integrity-authenticity) - [Meta: Our approach to labeling AI-generated content and manipulated media](https://about.fb.com/news/2024/04/metas-approach-to-labeling-ai-generated-content-and-manipulated-media/) ## Questions people ask ### Do UK advertising rules say I must label AI-generated ads? The ASA and CAP said in June 2025 that the CAP and BCAP Codes contain no specific rules on AI. What applies is the existing code: an ad must not materially mislead, however it was made. Separate from the code, platforms set their own label rules, such as TikTok's requirement to label realistic AI-generated content. ### Can I use an AI image of a celebrity in my ad? Avoid it. The ASA says an AI-generated celebrity endorsement that looks real enough to mislead is likely to breach the CAP Code, and implying an endorsement that does not exist is especially problematic. The ASA also warns of separate legal risks from using a public figure's likeness without permission, which it does not regulate itself. ### If the AI wrote the ad, am I still responsible? Yes. The ASA has said that even when campaigns are generated or distributed by automated methods, the advertiser keeps primary responsibility, citing its Stripe & Stare Ltd ruling. You also need evidence for any claim the ad makes. Treat AI as a drafting assistant and check every line before it goes out. ### Can I use AI images to show what my product does? Only if they show what it really does. The ASA says AI images used to make efficacy claims can mislead if they do not accurately reflect the product, in the same way as photoshopped or filtered images. Use AI images for mood or illustration, and show real results with real photos you can back up.