Disclosure of AI in advertising

Over the past few years, we have witnessed dramatic developments for AI with applications across many sectors that has been drawing the attention of regulators, now including the use of AI in advertising.

On 29 May 2025 the Committee of Advertising Practice (CAP) published some new guidance on their position regarding disclosure of artificial intelligence (AI) in advertising. This is particularly pertinent at the current time with Meta planning to help advertisers through new AI tools that will enable entire ads to be created using AI. Something which could have a big impact on traditional advertising creation and the advertising ecosystem.  

At present there is no express legal requirement to disclose the use of AI in ads within the UK and no rulings from the Advertising Standards Agency (ASA) to guide our thinking on this issue. However, it is important to be mindful of the Incorporated Society of British Advertisers (ISBA) and the Institute of Practitioners in Adverting’s (IPA) guidance published back in 2023. This outlines their guiding principles on using AI in advertising, of which CAP emphasised the principle of transparency in the use of AI where it features prominently and is unlikely to be obvious to consumers. Which raises the question, where ads contain AI generated content, should it be labelled as such?

In light of this, CAP advises that advertisers consider these two questions:

1. Is the audience likely to be misled if it is not disclosed?
2. If there is danger of being misled, is the disclosure clarifying the ad’s message or contradicting it?

This is useful guidance from CAP which builds on the principles set out in the CAP code, but it is important to remember that disclosure alone may not mitigate the harm caused by an otherwise misleading message.

So, if AI generated gives a misleading impression of the product you are advertising, this is unlikely to be mitigated by a simple disclosure. A good example of this would be the impact of cosmetics. If the AI generated images suggest an exaggerated impact on appearance then this would be misleading.

At the same time, there could be examples where a disclosure could help to clarify the situation for consumers and effectively mitigate any potential to mislead consumers. For example, if the advertising takes the form of an interactive AI lead experience or the product being advertised includes AI it may be appropriate to inform consumers in the ad itself. In this circumstance it may also be appropriate to consider whether it is necessary to inform consumers of the risk of hallucinations and whether the AI is generally compliant with other consumer protection regulation.

In some instances, no disclosure may be needed at all if the AI simply creates advertising content which accurately reflects the product.

We will certainly learn more in the coming months about how the ASA applies these principles through decisions. More generally it will be interesting to monitor the developments within the sector especially given the European Union’s AI Act which came into force in August last year. This seeks to categorise the use of AI based on the risk of harm. Although the use of generative AI is not classified as high-risk, it must adhere to the transparency requirements, one of which is to disclose where content was generated or modified by AI. This is something that all brands should keep on their radar, especially those advertising globally.

Whilst there is a lack of direct regulation, it is crucial to keep in mind that the same rules on advertising apply, regardless of whether AI is used or not. Principally the prohibition on misleading advertising, which may be at greater risk with the use of AI in the production of ads.