How to Label AI-Generated Product Images (2026 Rules)

A practical guide to which AI product visuals need disclosure, which are exempt, and how to build labeling into your image pipeline without slowing your catalog down.

|AI regulation e-commerce compliance AI product photography

As of August 2026, the EU AI Act's transparency obligations under Article 50 are enforceable — and they apply to a lot of e-commerce imagery. If you generate a lifestyle scene around a product, place a garment on an AI model, or synthesize a backdrop that never existed, that image is synthetic content under the regulation and needs to carry machine-readable provenance data.

The confusion in most brand teams right now isn't whether to comply. It's what counts. Background removal is not the same as background generation. Color correction is not the same as inventing a beach. The line the regulation draws is narrower than the panic suggests, and knowing exactly where it sits saves you from labeling 40,000 SKUs that never needed it.

This guide covers what triggers a labeling obligation for AI-generated product images, the two different types of disclosure the law requires, how the C2PA metadata standard actually works, and how to wire all of it into an existing catalog workflow.

What actually counts as an AI-generated product image

Article 50 targets content that is artificially generated or manipulated in a way that a reasonable viewer would take as a real depiction. That framing does most of the work. Editing a photograph is not generating one.

The practical test: did the AI invent visual information that was never in front of a camera, and would a shopper reasonably believe it was photographed?

Generally in scope

  • AI-generated lifestyle scenes and backdrops
  • AI models wearing your garment
  • Ghost mannequin effects synthesized from a flat lay
  • Products composited into rooms or environments that were never shot
  • Generated variant colorways presented as photographs

Generally out of scope

  • Background removal and white-background cutouts
  • Color correction and white balance
  • Dust, lint, and blemish removal
  • Upscaling, sharpening, denoising
  • Cropping, straightening, resizing

The second column is ordinary retouching. It has been standard practice since long before generative models, it does not fabricate a scene, and it does not mislead a buyer about what the product is. A cutout of a real shoe on white is still a photograph of that shoe.

The gray zone

Generative fill used to extend a real photo — widening a backdrop to hit a 1:1 crop, filling a corner of a seamless sweep — sits between the two columns. The safe read: if the fill is a continuation of a surface that genuinely existed in the frame, it behaves like retouching. If it adds objects, environments, or people, treat it as generated and label it.

The two obligations people keep conflating

Article 50 imposes two separate requirements, and they have very different scopes. Most of the confusion in brand Slack channels comes from treating them as one rule.

ObligationFormWho it applies toScope
Machine-readable markingEmbedded metadata, invisible to shoppersProviders of the AI systemEssentially all synthetic images
Visible disclosureOn-page label a shopper can readDeployers — you, the brandDeepfakes: realistic depictions of identifiable real people or events

The first is largely handled upstream. If you use a reputable AI imaging tool, the provenance marking should already be embedded in the file it hands you — your job is to verify it survives your pipeline, not to create it.

The second is where brands have real work to do, and it's narrower than most assume. A generated backdrop behind a coffee grinder does not depict an identifiable real person. An AI model rendered from scratch, who resembles no actual human, generally doesn't either. But if you generate imagery based on a real person's likeness — a founder, an athlete, a licensed ambassador — you are squarely in visible-disclosure territory.

Practical read

Most catalogs need metadata integrity across the board and visible labels on a small subset. Audit for the subset rather than stamping a disclaimer on every product page — over-labeling trains shoppers to ignore the label and can itself read as misleading.

How C2PA provenance metadata works

The machine-readable requirement is met in practice through C2PA — the Coalition for Content Provenance and Authenticity standard, backed by Adobe, Microsoft, Google, OpenAI, and the major camera manufacturers. It is the de facto answer to "how do I mark this in a detectable way."

A C2PA manifest is a cryptographically signed record attached to the image file. It records what created the image, whether a generative model was involved, and what edits were applied afterward. Because it's signed, tampering is detectable — you can't quietly strip the "AI-generated" assertion and keep a valid manifest.

The catch, and it's a serious one for e-commerce: most image pipelines destroy it.

Where provenance metadata typically gets lost
CDN re-encoding
Very common
WebP/AVIF conversion
Very common
Thumbnail generation
Common
Marketplace re-upload
Common
Social platform export
Common

Standard image optimization — the exact thing every e-commerce team does for page speed — strips non-essential metadata by default. A brand can generate a properly marked image, upload it, serve it through a CDN, and end up shipping a stripped file to every shopper. The obligation sits with the AI provider, but a pipeline that silently discards provenance undermines the whole regime and leaves you unable to prove what you did.

Fixing it is a configuration problem, not a rebuild: most image CDNs and build tools have a metadata-preservation flag that is simply off by default.

Building labeling into your catalog workflow

The efficient approach is to classify at generation time rather than audit at publish time. Once an image is sitting in your DAM alongside 30,000 others, working out how it was made is archaeology.

1. Tag at the source

Every image entering your asset library gets one of three flags: photographed, retouched, or generated. This is a single field, and it answers every compliance question you will be asked later. Tools that produce generated imagery should write this automatically — Retouchable, for instance, distinguishes retouching operations from generative ones in its job records, so the classification comes from the pipeline rather than someone's memory.

2. Preserve metadata through delivery

Audit each transform step in your delivery chain and enable metadata preservation. Test by pulling a live production image URL and inspecting it with a C2PA verification tool — do not assume the flag worked.

3. Identify your visible-disclosure subset

Filter for generated assets that depict identifiable real people. For most catalogs this is a small, knowable list: ambassador campaigns, founder imagery, licensed likenesses. Those get an on-page label.

4. Write the label like a human

Disclosure needs to be clear and accessible, not buried in a footer. It also doesn't need to be defensive.

WeakBetter
"Image may contain AI elements.""Model imagery generated with AI. Garment shown is the actual product."
"AI""Scene created with AI. Product photographed in studio."
Buried in terms of serviceCaption directly beneath the image

5. Keep records

Retain generation logs — model used, prompt or input, date, source photograph. If a regulator or marketplace asks how an image was produced, a timestamped record is the entire answer.

What this means beyond the EU

If you sell into the EU, the rules apply regardless of where you're headquartered. But the broader shift is not really regulatory — it's platform-driven, and it's moving faster than legislation.

Aug 2026EU transparency rules enforceable
3Major platforms auto-detecting C2PA
0Extra steps if tagged at generation

Major social and marketplace platforms already read provenance metadata and apply their own "AI info" labels automatically, without asking you. That means the label appears whether or not you were planning to disclose — and a brand that gets auto-labeled after saying nothing looks worse than one that disclosed up front.

Meanwhile, existing consumer protection law already covers the substantive risk. Any image that misrepresents the product — a color you don't sell, a fit that isn't achievable, an included accessory that isn't in the box — was actionable long before the AI Act existed. Generative tools make that failure mode easier to hit by accident, which is the real reason to keep generation confined to context rather than product.

The durable rule

Generate the scene, never the product. Backdrops, lighting environments, and staging can be synthetic. The item itself — its shape, color, texture, hardware, and proportions — must match what ships. That single constraint keeps you compliant under transparency rules, consumer protection law, and marketplace policy simultaneously.

A 30-minute compliance audit

You can get a realistic picture of your exposure in one sitting. Work through it in order.

StepWhat to checkTime
1List every tool in your image pipeline that can generate, not just edit5 min
2Estimate what share of live catalog images came from those tools10 min
3Pull three live product image URLs, inspect for a C2PA manifest5 min
4Identify any generated imagery depicting real, identifiable people5 min
5Confirm your DAM has a field distinguishing generated from retouched5 min

Step 3 is where most teams get a surprise: the manifest was there at upload and gone at delivery. Step 5 is the one that pays off longest — the brands that will find the next round of rules trivial are the ones that already know, per asset, how each image was made.

None of this requires abandoning AI imagery. Generated backdrops, on-model visuals, and synthetic staging remain the fastest way to build a catalog, and the transparency rules were written to make them trustworthy, not to make them unusable. The cost of compliance is a metadata flag and a database column — not a return to a rented studio.

Frequently Asked Questions

Do I need to label AI background removal on product photos?

No. Background removal is retouching, not generation — it removes information from a real photograph rather than inventing new visual content. The same applies to color correction, dust and blemish removal, upscaling, sharpening, and cropping. Labeling obligations attach to images where AI fabricated a scene, environment, or person that a shopper would reasonably take as photographed.

What is C2PA and do I have to implement it myself?

C2PA is a content provenance standard that embeds a cryptographically signed record in an image file describing how it was created and edited. The obligation to apply it sits with the AI provider, not the brand — reputable generative tools embed it automatically. Your responsibility is making sure your CDN, format conversion, and thumbnail steps do not strip that metadata before the image reaches shoppers, which they do by default in most setups.

Does an AI-generated model count as a deepfake requiring visible disclosure?

Generally no, if the model is entirely synthetic and does not resemble an identifiable real person. Visible disclosure under Article 50(4) targets realistic depictions of actual people or events. If you generate imagery based on a real person — a founder, athlete, or brand ambassador — that does require a visible on-page label, and it requires their consent independently of the AI rules.

What happens if my CDN strips the provenance metadata?

You end up serving unmarked synthetic images even though the tool that made them marked them correctly. The direct legal obligation rests with the AI provider, but you lose your ability to demonstrate how an image was produced, and platforms that auto-detect provenance will not apply their own AI label — which can look like concealment if the origin surfaces later. Most image CDNs have a metadata preservation setting that is off by default; turn it on and verify against a live production URL.

Do these rules apply if my brand is based outside the EU?

If you sell to customers in the EU, yes. The AI Act applies based on where the content is placed on the market, not where the company is headquartered. Beyond the EU, the practical pressure comes from platforms — several major marketplaces and social networks already read provenance metadata and apply their own AI labels automatically, regardless of jurisdiction.

Generate the scene. Never the product.

Retouchable keeps your actual product pixel-accurate while AI handles backdrops, staging, and on-model visuals — with every job recorded so you always know how an image was made.

Try Retouchable Free No credit card required