Skin-tone retouching: UK ad approval checks

UK brands should approve skin-tone retouching only after checking that it preserves the person’s appearance, does not exaggerate a product’s performance and works consistently across campaign formats. Compare the final creative with the original capture, examine any implied claims and review how lighting, grading and AI tools have affected different people—not just whether the image looks polished.

Why skin tone needs its own approval check

A request to “brighten the image” can become lighter skin. A warm lifestyle grade can make one person look healthy and another look unnaturally orange. Automated enhancements can alter faces unevenly within the same family photograph.

These are not simply production faults. They can change what an advert communicates about beauty, belonging and who a product is for.

Colourism means prejudice or unequal treatment based on skin shade, often favouring lighter skin. An edit can echo that hierarchy without anyone deliberately requesting it. Risk increases when lighter skin repeatedly receives the most flattering treatment, while darker skin loses detail or appears as a problem to correct.

There is no single multicultural British view of acceptable retouching. Expectations vary by generation, faith, region and individual, as well as by category and personal experience. Do not ask one colleague to approve an image on behalf of everyone who shares their background.

Separate compliance from cultural judgement

An image can be legally defensible but culturally clumsy. Equally, a culturally sensitive image can still mislead consumers about what a product does. Your approval process needs both checks.

Could the image mislead?

For non-broadcast advertising, the CAP Code’s misleading advertising rules are the starting point. Rule 3.1 prohibits materially misleading advertising. Rule 3.7 requires documentary evidence for objective claims capable of substantiation before publication.

Those claims need not appear in words. A beauty image can imply that a foundation produces a particular finish, a serum reduces visible pigmentation or a moisturiser creates a particular glow.

If editing supplies the result the product is supposed to deliver, stop and review the claim. Removing pigmentation in an advert for a product promoted to improve its appearance is materially different from removing a temporary mark in an unrelated furniture campaign.

Retouching is not automatically prohibited. The question is what consumers are likely to understand from the whole advert, and whether the presentation exaggerates performance. Do not assume a small “image retouched” disclaimer makes a misleading demonstration acceptable.

Could the treatment cause serious offence?

The CAP Code’s harm and offence rules prohibit advertising likely to cause serious or widespread offence, with particular care required around grounds including race, religion, sex and disability. Context, medium, audience and prevailing standards matter.

Not every poor colour grade reaches that threshold. Nevertheless, depicting darker skin as inherently dirty, undesirable or something to escape creates a clear reason to escalate the work.

Broadcast campaigns require assessment under the applicable BCAP rules and clearance process. Do not treat digital approval as clearance for television.

Define the intended appearance before the shoot

Start with a brief that describes lighting and finish, rather than an ethnic shorthand. “Premium”, “clean” and “aspirational” are not instructions to lighten skin.

For photography and video, agree:

• The lighting references and intended contrast, with examples across relevant skin tones.
• Which temporary features may be retouched and which characteristics should remain.
• Whether the product demonstration must remain untouched in particular areas.
• Who approves skin rendering, including any agreed consultation with talent.

Capture suitable reference material and use a colour-managed workflow. A grey card or colour chart can support technical consistency, but it does not decide whether a creative treatment is respectful or a product claim is truthful.

Keep the approved original files. Without a reliable comparison, reviewers are left debating memory and preference rather than identifying what changed.

Distinguish correction from alteration

White balance, exposure correction and skin-tone alteration can overlap. A correction may legitimately change the recorded colour of skin because the camera captured a colour cast. Conversely, an edit labelled “exposure only” can still erase someone’s natural depth of tone.

Ask three practical questions:

1. What caused the change? Was it correcting capture conditions, matching shots or deliberately changing the person’s appearance?
2. What else changed? Have skin texture, facial features or the contrast between people shifted?
3. What does the finished advert imply? Does the person now appear lighter because of a product, a transformation or a supposed improvement in status?

Do not impose a simplistic rule that every pixel must match the source. Instead, require a documented rationale for material changes and a final image that remains faithful to the person and the claim.

Pay particular attention to group shots. Applying one global adjustment may leave a lighter-skinned person well exposed while flattening detail on a darker-skinned person. Equal settings do not guarantee equal quality.

Treat AI enhancement as a fresh creative version

Do not assume an AI-assisted upscale, relight or background replacement leaves the subject untouched. Depending on the tool and settings, processing may change skin texture, facial details, hair edges or apparent skin shade.

A background change can also alter perceived skin tone through surrounding colour and contrast, even when the face itself has not been intentionally edited.

Review the exported result, not just the editing instructions. Compare faces at a useful viewing size and inspect the finished placement at normal audience scale.

For a synthetic person, there may be no original photograph to check. Review the prompt, selection process and final image instead: are people consistently made lighter, younger or more conventionally attractive when the brief asks for “successful” or “professional”?

AI-generated imagery does not remove the advertiser’s responsibility for the message. Nor should a synthetic face be presented as evidence of a real person’s product experience.

Use this pre-launch approval checklist

Give one named owner responsibility for completing these checks and recording unresolved issues.

• Source comparison: Review the original capture, approved master and final export together.
• Skin rendering: Check for unwanted lightening, darkening, colour casts and lost detail across everyone pictured.
• Claim integrity: Identify what the image says about product performance, including implied results.
• Evidence: Confirm that objective claims have appropriate substantiation and that editing has not manufactured the demonstrated effect.
• AI changes: Inspect enhanced or generated assets for unintended changes to appearance.
• Sequence: Check before-and-after layouts, transitions and surrounding copy for a darker-to-lighter “improvement” story.
• Placements: Review social crops, thumbnails, retailer assets and video grades—not only the hero image.
• Sign-off: Record the approved version, decision-maker and any required fixes; reopen approval after material changes.

Use a calibrated reference display where available, alongside representative consumer devices. You cannot control every screen, but you can avoid approving solely from a compressed screenshot in a messaging thread.

When to stop rather than tweak

Pause publication if the team cannot explain a material skin-tone change, if an edit creates the advertised result, or if the concept depends on darker skin representing an inferior “before”. These are not issues to solve with a warmer caption.

For less serious production faults, request a targeted correction and repeat the comparison. Avoid sending the vague instruction “make everyone look more natural” without defining the problem: it invites another round of subjective editing.

Review cultural risk before launch

Review the complete advert—copy, image, sequence, sound and placement—with people who understand the relevant audiences and can challenge the creative rationale. Ask what viewers might infer about skin shade, attractiveness and belonging, while recognising disagreement within communities.

Kulture adapts ad copy, images and video for UK communities and runs cultural risk reviews. Build that review in before final delivery, while the team still has time to change the treatment rather than defend it.