Two ASA rulings found undisclosed beauty filters misleading

The persona

This entry addresses no single persona but the closest verified UK regulatory precedent on a question this site tracks across many personas: whether an ad's digitally altered visual element must be disclosed. No ruling naming a virtual influencer specifically could be verified here; the nearest matching precedent concerns Instagram beauty filters applied to real people in cosmetic-product ads, a narrower but directly comparable disclosure question.

What the documents establish

On 3 February 2021 the UK Advertising Standards Authority published two related rulings, both upheld. In Skinny Tan Ltd in association with Elly Norris, the regulator found that an Instagram Story reposted onto the brand's own account, showing influencer Elly Norris applying a tanning product while an in-app 'Perfect Tan' filter altered her appearance, was misleading because it exaggerated the product's likely effect. In We Are Luxe Ltd, with Cinzia Baylis-Zullo, decided the same day, the ASA reached the same conclusion about a separate filter that added freckles and altered skin tone in a product demonstration, despite an '#ad' tag on the post. Both rulings required only that the ads not reappear in that form and that the brands not apply filters likely to exaggerate a product's effect. A later ASA/CAP guidance page, retrieved 16 September 2026, cites both rulings as its example of how social media filters are treated under the same framework as retouching.

Character versus company

In both rulings the influencers said they had not intended to mislead, and in one case that the filter was unrelated to the product's advertised effect; the ASA's decision rested on the brands' own choice to repost that content onto their commercial accounts, not on the influencers' private intent. That is the pattern this site tracks elsewhere: an operator's decision to publish or amplify content is what a regulator evaluates, separate from the sincerity of the person whose image appears in it. Neither ruling addresses a fully computer-generated character, and the gap between a real person under a filter and a persona that is entirely CGI is not closed by these documents.

What to watch

No ruling reviewed here names a virtual influencer or fully CGI persona directly; treat these two filter rulings as the closest available analogue, not an equivalent. It is worth watching whether the ASA rules on a virtual-influencer campaign specifically, showing how it applies this framework to a persona with no real person behind it at all. This is an editorial framing of an open gap, not a prediction of the outcome.

• Has the ASA since ruled on an ad featuring a fully computer-generated persona, rather than a filtered real person?
• What CAP Code rules would apply differently to an entirely CGI character than to a filtered human influencer?
• Does an '#ad' tag change the ASA's analysis of an undisclosed filter, or are the two disclosures treated separately?

The two February 2021 rulings are a real, upheld precedent on undisclosed digital alteration in UK advertising, but they are a precedent about filtered humans, not CGI personas, a gap worth stating plainly rather than papering over with the nearest available case.