Amazon’s AI Image Disclosure Rule 2026: How Sellers Must Label AI-Generated People in Listings & A+ Content

‍

AI-generated lifestyle imagery has quickly become part of the Amazon creative workflow.

Brands can now create lifestyle scenes, models, demonstrations, backgrounds, and campaign concepts faster than traditional photoshoots ever allowed.

But Amazon has added an important new compliance step.

Amazon says sellers must now tag product media containing photorealistic AI-generated people with specific metadata before uploading it to product listings and A+ Content in its worldwide stores. The required keyword is:

contains-synthetic-performer

Amazon instructs sellers to add that value to the dc:subject XMP metadata field using an IPTC-compatible metadata editor. Where applicable, Amazon may then display an indicator informing shoppers that the media contains an AI-generated person. (Amazon Seller Central)

This is an important distinction:

Amazon is not banning AI-generated people. It is requiring disclosure when photorealistic people themselves are AI-generated.

For sellers and creative agencies, that changes the workflow from:

Create → Export → Upload

to:

Create → Verify → Tag → Validate → Upload

At Adorbix, we see this as more than a technical metadata requirement. It is now part of Amazon creative compliance.

‍

What Exactly Did Amazon Change?

‍

Amazon's announcement says product images containing photorealistic AI-generated people must carry the required metadata before being added to listings or A+ Content.

Amazon says the requirement applies across its worldwide stores. (Amazon Seller Central)

The required process is straightforward:

1. Create or finalize the media.
2. Open the final file with an IPTC-compatible metadata editor.
3. Add:

contains-synthetic-performer

to the:

dc:subject (XMP) field.

4. Save the final media.
5. Verify the metadata remains present.
6. Upload the asset to Amazon.

Amazon says that when this metadata is included, it can add an indicator for customers identifying media that contains AI-generated people, where applicable. (Amazon Seller Central)

‍

What Counts as a Photorealistic AI-Generated Person?

‍

The key question is not:

“Was AI used anywhere in the image?”

The question is:

“Does the image or video contain a photorealistic person that was generated by AI?”

Think of examples such as:

• AI-generated lifestyle models
• AI-generated fitness models
• Synthetic people wearing apparel
• AI-generated hands or bodies presented as realistic humans
• AI-generated people demonstrating products
• Realistic synthetic spokesperson-style people

If the creative depicts a realistic human who does not actually exist and was generated synthetically, it falls squarely within the kind of content Amazon's disclosure rule is designed to address.

‍

What Does NOT Require the Metadata?

‍

Amazon explicitly lists several exemptions.

You do not need this metadata when the media:

• Contains only real people, even if those real people were modified using AI tools
• Contains only characters from movies, television, games, documentaries, streaming content, or other expressive works
• Contains no people
• Contains people who are not photorealistic (Amazon Seller Central)

This makes the rule much narrower than:

“All AI images must be labeled.”

They do not.

‍

Examples: Tag or No Tag?

‍

CreativeTag Required?AI-generated realistic female model wearing leggingsYesAI-generated realistic man holding your productYesCompletely AI-generated kitchen with no peopleNoAI-generated warehouse background with no humansNoReal model with AI color correctionNoReal model with background replaced using AINo, based on Amazon's stated exemption for real people altered with AICartoon personNoStylized illustration of a personNo, if not photorealisticStandard product render with no peopleNo

The distinction Amazon has drawn is unusually specific:

Synthetic photorealistic person → disclose.

AI used elsewhere → not automatically subject to this particular metadata requirement.

(Amazon Seller Central)

‍

Why Amazon Introduced the Rule

‍

Amazon says recent legislation requires disclosure when advertising images or videos contain photorealistic AI-generated people. Its new metadata workflow is intended to enable that disclosure across marketplace media. (Amazon Seller Central)

There is a broader industry reason behind this.

AI-generated people are becoming increasingly difficult for consumers to distinguish from real photography.

A customer viewing:

• A fitness model wearing leggings
• A parent holding a baby product
• A stylist using beauty equipment
• A chef using cookware

may reasonably assume that they are seeing an actual person photographed with the real product.

The new rule introduces a transparency layer without prohibiting brands from using generative AI.

‍

This Is Particularly Important for Amazon Lifestyle Images

AI is especially attractive for secondary product imagery because lifestyle production can be expensive.

Traditionally, creating one lifestyle photo might require:

• Model
• Photographer
• Studio
• Location
• Lighting
• Stylist
• Retouching
• Product samples

Generative AI can reduce that production burden dramatically.

But if the lifestyle model is synthetic and photorealistic, the compliance workflow now needs to include metadata disclosure before the image reaches Amazon.

That affects categories such as:

• Apparel
• Fitness
• Beauty
• Personal care
• Home
• Baby products
• Medical accessories
• Sports
• Jewelry
• Footwear

Any category where human use is part of the merchandising experience should pay particular attention.

‍

The Rule Also Applies to A+ Content

‍

This is easy to miss.

The requirement is not limited to the normal image gallery.

Amazon specifically says sellers must tag applicable media before adding it to:

Listings

and

A+ Content

across Amazon stores worldwide. (Amazon Seller Central)

That matters because A+ Content frequently contains the most lifestyle-heavy creative on a listing.

For example:

Hero banners

AI model holding the product.

Lifestyle modules

Synthetic family using the product.

Feature modules

AI athlete demonstrating activewear.

Brand story imagery

Photorealistic synthetic founders, employees, customers, or lifestyle models.

If those people are AI-generated and realistic, the asset workflow now needs the synthetic-performer tag.

‍

What About AI-Generated Backgrounds?

‍

This is one of the most practical questions for creative teams.

Suppose you photograph a real product and place it into an AI-generated:

• Bathroom
• Gym
• Bedroom
• Kitchen
• Warehouse
• Beach
• Spa

but there are no AI-generated people.

Amazon's announced rule says disclosure is required specifically for photorealistic AI-generated people, and it explicitly says media with no people does not require the metadata. (Amazon Seller Central)

So the disclosure rule should not be interpreted as a universal tag for every AI-generated background.

‍

What If a Real Model Is Enhanced With AI?

‍

Amazon explicitly says the metadata isn't required if the media contains only real people—even when those people have been altered using AI tools. (Amazon Seller Central)

That distinction matters for standard post-production.

Examples may include:

• Skin retouching
• Lighting correction
• Background replacement
• Clothing cleanup
• Color correction
• Minor generative retouching

The important factor is that the actual person represented is real rather than a completely synthetic photorealistic performer.

‍

Amazon Is Not Saying You Can Misrepresent the Product

‍

This rule should not be confused with permission to use AI however you want.

Even when the synthetic-performer metadata is present, your creative should still represent the product accurately.

A disclosure tag does not make it acceptable to generate:

• The wrong product color
• Fake accessories
• Extra product features
• Incorrect dimensions
• Altered packaging
• False logos
• Different materials
• Impossible product performance

The disclosure is about who the person is.

It doesn't eliminate the seller's responsibility to keep the product itself accurate.

For Amazon sellers, this is especially important because AI can easily make a product look slightly different while generating lifestyle scenes.

‍

The Hidden Risk: Metadata Can Disappear During Production

‍

Creative teams should also update their file-handling workflow.

Metadata added to one version of an image is not useful if the final exported file loses it.

Some workflows may strip metadata when files are:

• Re-exported
• Compressed
• Optimized
• Converted between formats
• Passed through certain design applications
• Reprocessed by external tools

Therefore, the correct workflow isn't:

Tag the Photoshop master and assume you're done.

It should be:

Tag and verify the exact file that will be uploaded to Amazon.

This matters especially for agencies working with several designers or multiple rounds of client revisions.

‍

The New Amazon AI Creative Workflow

‍

For brands producing AI lifestyle assets at scale, use a structured workflow.

Step 1 — Classify the Creative

Ask:

Does the final media contain a person?

If no:

This specific disclosure requirement does not apply.

If yes:

Continue.

Step 2 — Determine Whether the Person Is Real or Synthetic

Real person

Amazon says no synthetic-performer metadata is required, even if AI tools were used to alter the media. (Amazon Seller Central)

AI-generated person

Continue.

Step 3 — Determine Whether the Person Is Photorealistic

Cartoon / stylized / clearly non-photorealistic

Amazon says the disclosure isn't required under this rule.

Photorealistic synthetic person

Tag it.

Step 4 — Add the Required Metadata

Add the exact keyword:

contains-synthetic-performer

to:

dc:subject (XMP)

using an IPTC-compatible metadata editor. (Amazon Seller Central)

Step 5 — Perform Product Accuracy QA

Check:

• Product shape
• Color
• Texture
• Dimensions
• Logo
• Label
• Text
• Accessories
• Packaging
• Product count

AI disclosure and product accuracy are two separate quality checks.

Do both.

Step 6 — Export the Final File

Export at Amazon's required dimensions and quality.

Then verify the metadata after export.

Step 7 — Upload to Amazon

Use the validated asset in the relevant:

• Listing image gallery
• A+ Content
• Applicable video workflow

Amazon says the metadata allows it to add a customer-facing indicator where applicable. (Amazon Seller Central)

‍

What Sellers Should Audit Right Now

‍

If your creative team uses AI, perform a content inventory.

Start with your highest-revenue ASINs.

Check:

‍

Listing images

Does any secondary image use a synthetic photorealistic model?

‍

A+ Content

Do any modules contain AI-generated people?

‍

Videos

Are synthetic photorealistic performers used?

‍

New creative waiting for upload

Has metadata already been added?

‍

Agency files

Does your outside agency know about the requirement?

This shouldn't become a manual scramble every time an image goes live.

Build it into the creative SOP.

‍

The Adorbix AI Creative Compliance Framework

‍

At Adorbix, AI-assisted Amazon creative shouldn't be treated as:

Generate image → Looks good → Upload.

We recommend five checkpoints.

‍

01 — PRODUCT LOCK

Confirm:

• Exact product
• Correct packaging
• Correct color
• Correct material
• Correct proportions
• Correct branding

‍

02 — HUMAN CLASSIFICATION

Determine:

• Real person?
• Synthetic person?
• Photorealistic?
• Disclosure required?

‍

03 — METADATA

Where required:

contains-synthetic-performer

in the required metadata field.

‍

04 — AMAZON QA

Review:

• Listing policy
• Image quality
• Product claims
• A+ compatibility
• Mobile readability

‍

05 — FINAL VALIDATION

Verify the final upload asset—not merely the original working file.

The objective is simple:

AI speed without sacrificing Amazon compliance or product accuracy.

‍

How Adorbix Can Help Brands Adapt

‍

This policy is especially relevant to brands producing large volumes of Amazon creative.

At Adorbix, our workflow can connect:

‍

AI Lifestyle Creative

Generate scalable lifestyle scenes while maintaining product fidelity.

‍

Product Image Production

Keep:

• Product design
• Labels
• Colors
• Text
• Packaging

accurate.

‍

A+ Content

Build high-converting A+ assets while reviewing applicable AI-person disclosure requirements.

‍

Creative Compliance

Classify final images and videos based on whether they contain photorealistic AI-generated people.

‍

Metadata QA

Add and verify required disclosure metadata where applicable.

‍

Conversion Optimization

Compliance shouldn't mean boring creative.

The goal is still:

Better imagery → clearer benefits → stronger customer confidence → higher conversion.

‍

Why This Rule Could Actually Improve AI Creative

‍

Disclosure requirements may sound like another compliance burden.

But they can encourage brands to become more disciplined.

Before using an AI model, ask:

Does this person actually improve the customer's understanding of the product?

If not, you may not need a human model at all.

A strong product image could use:

• Product macro
• Feature demonstration
• Environment
• Dimensions
• Infographic
• Product comparison

instead.

AI should improve communication—not become decoration.

‍

10 Mistakes Sellers Should Avoid

‍

1. Assuming every AI-generated image requires the tag

It doesn't. Amazon's rule specifically concerns photorealistic AI-generated people. (Amazon Seller Central)

‍

2. Forgetting A+ Content

The rule applies there too.

‍

3. Tagging the working file but uploading another export

Verify the final asset.

‍

4. Assuming an AI background automatically requires disclosure

No person means this particular synthetic-performer metadata isn't required. (Amazon Seller Central)

‍

5. Confusing AI retouching with synthetic people

Amazon explicitly exempts media containing only real people even when AI altered the media. (Amazon Seller Central)

‍

6. Using the disclosure to justify inaccurate products

The product still needs to be represented truthfully.

‍

7. Letting different designers follow different rules

Create one shared SOP.

‍

8. Forgetting video assets

Amazon's announcement discusses disclosure for applicable image/video files, so video workflows should also be reviewed. (Amazon Seller Central)

‍

9. Assuming the visible customer indicator is something sellers manually design

Amazon says it can add an indicator based on the metadata where applicable. (Amazon Seller Central)

‍

10. Waiting until upload day to check compliance

Metadata should be part of creative production, not an afterthought.

‍

Amazon AI Media Decision Matrix

Media TypeSynthetic Performer Tag?Real product onlyNoAI-generated room, no peopleNoReal human photographed normallyNoReal human edited using AINo under this ruleCartoon AI characterNoPhotorealistic AI lifestyle modelYesPhotorealistic synthetic spokespersonYesAI-generated realistic person in A+ bannerYes

This matrix is based on the scope Amazon announced for photorealistic AI-generated people and its stated exemptions. (Amazon Seller Central)

‍

Amazon AI Creative Checklist for 2026

Before uploading creative:

• Does it contain a person?
• Is the person real or synthetic?
• Is the synthetic person photorealistic?
• Is contains-synthetic-performer required?
• Was it added to dc:subject XMP?
• Has the final exported file been verified?
• Is the product 100% accurate?
• Are logos correct?
• Is packaging correct?
• Are colors accurate?
• Are product claims supportable?
• Does the asset meet Amazon image requirements?
• Has A+ Content been reviewed separately?
• Have video assets been checked?
• Has the team documented the final compliant file?

‍

FAQ

‍

Does Amazon allow AI-generated people in product images?

Amazon's announcement does not prohibit them. Instead, it requires sellers to tag media containing photorealistic AI-generated people with specific metadata before adding it to listings and A+ Content. (Amazon Seller Central)

‍

What metadata does Amazon require?

Amazon instructs sellers to add:

‍

contains-synthetic-performer

to the:

‍

dc:subject XMP field

using an IPTC-compatible metadata editor. (Amazon Seller Central)

‍

Does this apply to A+ Content?

Yes. Amazon explicitly names both listings and A+ Content in the requirement. (Amazon Seller Central)

‍

Does an AI-generated background require disclosure?

Not under this specific requirement if the media contains no people. Amazon explicitly says media with no people doesn't require the synthetic-performer metadata. (Amazon Seller Central)

‍

What if I use a real person and edit them with AI?

Amazon says you do not need the metadata if the media contains only real people, even if those people were altered with AI tools. (Amazon Seller Central)

‍

What about cartoon or obviously stylized AI people?

Amazon says the requirement does not apply when media does not contain photorealistic people. (Amazon Seller Central)

‍

Will shoppers see an AI label?

Amazon says that when the required metadata is included, it will add an indicator informing customers that the image contains AI-generated people where applicable. (Amazon Seller Central)

‍

Does this apply worldwide?

Amazon says the requirement applies before adding applicable media to listings and A+ Content in any of its worldwide stores. (Amazon Seller Central)

‍

Do sellers need to change every old listing immediately?

Amazon's public announcement reviewed here clearly says sellers must now tag applicable media before adding it to listings and A+ Content, but it does not clearly set out a blanket retroactive remediation deadline for every previously published asset. For older creative, check your current Seller Central guidance and prioritize anything being edited or re-uploaded. (Amazon Seller Central)

‍

Can Adorbix help with Amazon AI creative compliance?

Yes. Adorbix can integrate AI image generation, product-accuracy QA, A+ Content production, disclosure classification, metadata checks, and Amazon-ready creative delivery into one workflow.

‍

Key Takeaways

‍

• Amazon now requires disclosure metadata for media containing photorealistic AI-generated people before it is added to listings or A+ Content. (Amazon Seller Central)
• The required keyword is contains-synthetic-performer.
• Amazon says it belongs in the dc:subject XMP field using an IPTC-compatible metadata editor. (Amazon Seller Central)
• The rule applies across Amazon's worldwide stores.
• AI-generated backgrounds without people do not fall under this specific disclosure rule.
• Real people altered using AI are expressly exempt from this particular metadata requirement. (Amazon Seller Central)
• Non-photorealistic people are also exempt.
• The rule applies to A+ Content, not just standard listing images.
• Brands should verify metadata on the final upload file.
• AI disclosure does not excuse inaccurate product representation.
• Creative teams should add compliance review to their normal production SOP.
• AI should increase creative speed without reducing transparency or product accuracy.

‍

Final Takeaway: AI Creative Is Still Powerful—But the Workflow Has Changed

‍

Amazon hasn't told sellers:

“Stop using AI.”

It has effectively told them:

“If you're showing shoppers a realistic person who was generated by AI, disclose it correctly.”

That's a meaningful but manageable change.

The old creative workflow was:

Generate → Design → Upload

The 2026 workflow should be:

GENERATE → VERIFY → DISCLOSE → VALIDATE → UPLOAD

At Adorbix, we believe that is exactly how professional Amazon AI creative should work.

Use AI to:

• Produce faster
• Explore more concepts
• Create better lifestyle storytelling
• Reduce production bottlenecks
• Scale A+ Content

But protect:

• Product accuracy
• Customer trust
• Brand consistency
• Amazon compliance

Because the goal isn't to make creative that merely looks real.

It's to make creative that sells the real product accurately, transparently, and convincingly.

‍