Quick Answer
Most major commerce and advertising channels permit at least some AI-assisted product imagery, but there is no universal rule. Walmart expressly allows accurate, truthful, rights-compliant AI imagery; Etsy generally requires original photos of the actual finished product; and TikTok Shop prohibits placeholders and digital renderings in product listings. Disclosure is a separate question: Google Merchant Center requires machine-readable metadata for generative-AI images, TikTok requires labels for substantially AI-generated promotional content, and regional laws can add visible-label duties.
This article is general information, not legal advice. For copyright, likeness, advertising, and disclosure issues, see the AI product photography legal guide and consult a lawyer about decisions that affect your business.
Permission, listing accuracy, and AI disclosure are separate tests
AI product-image compliance has three layers: whether the channel accepts the image, whether the image accurately represents the item, and whether AI use must be disclosed. A seller can satisfy one layer and still fail another. Adding an AI label, for example, does not make a synthetic main image acceptable on a marketplace that requires a real photograph.
- Permission: Does the channel permit a generated image, an AI edit of a real photo, or only a conventional photograph?
- Accuracy: Does the image show the product, variation, condition, scale, and included accessories the buyer will receive?
- Disclosure: Must the seller preserve metadata, select a label, or add visible text?
Platform policies also distinguish listing images from ads, organic posts, books, and creator videos. Amazon KDP's AI disclosure rule does not automatically apply to an Amazon Seller Central product listing. TikTok Shop's rules for product-listing images differ from its rules for AI-generated promotional content. Treat the surface and the image slot as part of the rule.
Can I use AI-generated product images on each major platform?
The safest cross-platform answer is to begin with a real product photograph, keep the product materially unchanged, and check the exact image slot before publishing. The current public rules differ enough that a blanket “AI images are allowed everywhere” claim would be misleading.
| Platform or surface | Current rule for product imagery | AI disclosure | Practical decision |
|---|---|---|---|
| Amazon Seller Central | Amazon's product-image rules require accurate product representation and impose category-specific main-image standards. Amazon also offers generative imagery for advertising, but that does not override listing-image rules. | The current Seller Central image guide does not state a general AI-disclosure step. Amazon KDP separately requires disclosure of AI-generated book content. | Use a faithful product image and check the category's main-image rules. Do not import KDP's disclosure rule into ordinary listings. |
| eBay | eBay permits its own AI background tool, but its picture policy bars inaccurate images and stock photos for used, damaged, or defective items. | No general AI-image disclosure step appears in the current picture policy. | For used or flawed goods, show the actual unit and its condition. |
| Google Merchant Center | Google's AI-generated content policy allows generative-AI images in the main, additional, and lifestyle image attributes. | Generative-AI images must retain the applicable IPTC DigitalSourceType metadata. | Confirm that export, CDN processing, and feed ingestion do not strip the required metadata. |
| Walmart Marketplace US | Walmart's US image guide states that AI-generated content must be accurate, truthful, rights-compliant, and consistent with Marketplace policies. It also requires actual product images that accurately reflect the item. | The current US image guide does not state a separate AI-disclosure step. | AI can change presentation, but it must not invent or alter the product being sold. See the Walmart AI-image policy guide and Walmart image-requirements guide. |
| TikTok Shop US listings | TikTok Shop's Product Listing Policy prohibits placeholders and digital renderings, requires accurate product images, and requires the main image to show the physical product on white. | The listing policy does not substitute disclosure for the ban on digital renderings. Separate promotional-content rules require disclosure when content is fully or significantly generated or edited with AI. | Use a real product photograph for the listing. Apply the AIGC label rules separately to shoppable videos, LIVE content, and other promotion. |
| Etsy | Etsy's Listing Image Requirements generally require the seller's own photographs of the actual finished product, not artistic renderings or stock photos. Limited mockup exceptions apply to certain customized items and production-partner products. | Seller-prompted AI creations sold as the item must disclose AI use in the listing description. | Do not replace a handmade physical item's required real photograph with a synthetic product render. |
| Pinterest's AI guidance says its Community Guidelines apply regardless of how a Pin was created, and Pinterest offers AI background generation for advertisers. | Pinterest adds an AI label when metadata, classifiers, or the content owner's declaration indicate AI generation or modification. | Preserve provenance signals and expect Pinterest to apply the label when its systems detect AI use. | |
| Meta ads | Meta allows generative-AI ad creative subject to its advertising rules. | Meta labels images or videos created or significantly edited with its own tools and, from its June 2026 rollout, can detect and label third-party AI creative through industry signals. | Accuracy still controls ad approval; an informational AI label does not cure a misleading product depiction. |
| Google Ads | Google provides generative image tools and holds generated creative to the same advertising policies, including misrepresentation rules. | Google added an AI-label setting in July 2026. Labels are available globally in “How this ad was made,” with visible overlays in the EU, India, and New York for assets designated as AI-created or edited. Election-ad disclosure remains a separate rule. | Use the AI-label control where the campaign's geography or applicable law requires it. |
| Shopify storefronts | Shopify is commerce infrastructure, not the seller of record. Its terms make the merchant responsible for store materials, required disclosures, consumer-protection law, and third-party rights. | Shopify's core terms do not create a universal AI-photo label for storefront product media. | Follow the laws and downstream channel rules that apply to the shoppers and ads you target. The Etsy and Shopify disclosure guide separates their rules. |
| Wayfair suppliers | Wayfair's public Supplier Code requires truthful, accurate product information and rights to all submitted imagery. It does not state a specific AI-image rule. | No AI-specific supplier disclosure appears in the public Supplier Code. | Do not infer permission from Wayfair's consumer AI tools; obtain current partner guidance for synthetic supplier imagery. See the Wayfair AI-image policy guide. |
| Faire brands | Faire's public photography guidance calls for clear, accurate product photographs, and its Brand Terms require high-resolution photographs. Neither publishes a specific generative-AI exception. | No AI-specific disclosure step appears in the public photography guidance. | Treat fully synthetic product replacements cautiously and ask Faire if the image is not a photograph of the real item. See the Faire AI-image policy guide. |
Policies change and can vary by country, category, campaign type, and image position. The table reflects public US-facing or global English-language sources reviewed on September 3, 2026, not private seller guidance or a guarantee of approval.
Which marketplaces restrict synthetic product-listing images?
Etsy and TikTok Shop publish the clearest restrictions on synthetic listing images, while eBay imposes an actual-item rule for used, damaged, or defective goods. These are image-source or image-type restrictions, not merely accuracy reminders.
What does TikTok Shop allow in product listings and promotional content?
TikTok Shop US does not allow placeholders or digital renderings in product-listing images. Its Product Listing Policy states, “Placeholders and digital renderings of a product are not allowed.” The same policy requires the main image to show the front physical view of the product on a pure white background; all listing images must accurately show what the customer receives.
TikTok Shop's separate AI-Generated Content Restrictions and Requirements permits responsible AI editing and third-party AI tools in promotional content when the result is truthful. Creators and sellers must proactively disclose content that is fully generated or significantly edited with AI, using visible text or TikTok's AI-generated content setting. Enabling the label does not, by itself, demote compliant content.
The help-desk guide to TikTok Shop and Instagram product-image rules covers the two surfaces separately.
What AI product images does Etsy allow?
Etsy distinguishes the item being sold from the photographs used to represent it. Seller-prompted AI artwork can qualify as a “Designed by a seller” item when the seller discloses AI use in the listing description. A physical product listing, however, generally needs the seller's own photos of the actual finished item.
Etsy permits computer-generated mockups only in limited cases. For personalized products, additional images may show customization options, but the first image must show a finished customized item similar to what the buyer will receive. For an original design printed by a production partner, a stock-photo mockup can illustrate the end product. These exceptions do not create blanket permission to replace any physical product photo with an AI rendering.
For a listing-form decision, use the Etsy and Shopify AI-disclosure guide.
When does eBay require a photograph of the actual item?
eBay requires an accurate image and prohibits stock photos for used, damaged, or defective items. That rule makes the actual unit's condition visible, which a synthetic replacement cannot establish. eBay's picture policy also requires at least 500 pixels on the longest side and bars added borders, promotional text, and watermarks.
AI editing is not categorically banned: eBay offers an AI background-removal and background-replacement tool. The edited result still has to represent the listed item and, for used or flawed goods, preserve the condition a buyer needs to inspect. See eBay's AI product-photo rules and the full eBay product-image requirements.
Which platforms require AI labels or metadata?
Google Merchant Center has a direct machine-readable requirement, TikTok Shop requires disclosure for significantly AI-generated promotional content, and Pinterest and Meta apply informational labels when their systems detect or receive AI signals. Google Ads now provides a broader labeling control for legal requirements that vary by geography.
What metadata does Google Merchant Center require for AI images?
Google Merchant Center requires images created with generative AI to retain IPTC DigitalSourceType metadata. Google's current policy lists TrainedAlgorithmicMedia, CompositeSynthetic, and AlgorithmicMedia as digital-source values to preserve, depending on how the image was created. The rule covers image_link, additional_image_link, and lifestyle_image_link attributes.
The operational risk is metadata loss. An image editor, CDN, storefront optimizer, or feed transformation can re-encode a file and remove embedded fields. Verify the final file served from the feed URL rather than checking only the local export. The Google Shopping AI-image guide and Google Shopping image requirements cover the rest of the feed workflow.
How do Pinterest and Meta label AI-generated images?
Pinterest and Meta use technical signals and platform data to apply labels; neither platform's public explanation says every minor AI edit receives the same treatment. Pinterest shows “AI modified” on image Pins when metadata, classifiers, or an owner declaration indicate AI use. For ads, the disclosure appears with other ad details.
Meta's current ads-transparency explanation says significant edits made with Meta's generative tools receive an “AI info” label. Meta's June 2026 update also began rolling out automatic detection of third-party AI creative through industry-standard signals, with the label available through “About this ad.” For organic Facebook and Instagram posts, Meta likewise labels detected or self-disclosed AI-generated content and places labels for minor AI edits in the post menu.
Pinterest's process is covered in the AI-generated Pin policy guide; paid campaigns are compared in the guide to AI product images in Meta and Google ads.
What changed for Google Ads AI disclosure in July 2026?
Google Ads now lets advertisers designate image and video assets as AI-created or AI-edited. The resulting disclosure appears globally in the “How this ad was made” section of My Ad Center; visible overlays appear in the EU, India, and New York for designated assets because Google identifies those regions as having applicable AI-transparency requirements.
This is a material change from the earlier election-only framing. Election advertisers still have a separate requirement to declare certain altered or synthetic content, but ordinary advertisers now also have AI-label controls. Google's own guidance says using the setting does not guarantee legal compliance, so campaign geography and the specific asset remain part of the decision.
For campaign-level steps, see the help-desk comparison of Meta and Google Ads AI-image rules.
Does the EU AI Act require every AI product photo to carry a visible label?
The EU AI Act does not automatically require a visible seller label on every AI-assisted product photo. Article 50 separates obligations for AI-system providers from obligations for businesses deploying deepfake content, and its standard-editing exception also matters.
From August 2, 2026, providers of systems that generate synthetic image, audio, video, or text content generally must make outputs machine-readable and detectable as artificial. That provider obligation does not apply when the system performs an assistive function for standard editing or does not substantially alter the input or its meaning. A limited grace period until December 2, 2026 applies only to the Article 50(2) marking duty for systems placed on the market before August 2, 2026.
A seller or advertiser acting as a deployer must visibly disclose an image when it constitutes a deepfake. The Act defines that category around content that appreciably resembles existing people, objects, places, entities, or events and would falsely appear authentic or truthful. The European Commission's Article 50 guidance says the assessment depends on resemblance, message, deployment context, and audience expectations. A product image that invents a realistic scene or alters the product may therefore require analysis; a routine crop or standard edit is not automatically in the same category.
National authorities enforce Article 50, and fines for covered violations can reach EUR 15 million or 3% of worldwide annual turnover. Because the result turns on facts and context, businesses targeting EU buyers should obtain legal advice rather than attaching or omitting a label based on a one-line rule.
Does Amazon require AI disclosure on product-listing images?
Amazon's current Seller Central product-image guide does not state a general AI-disclosure step for standard product-listing images. Amazon's explicit AI-content disclosure rule applies to Kindle Direct Publishing, where publishers must tell Amazon about AI-generated text, images, or translations.
The distinction matters because Amazon has several separate publishing and advertising surfaces. Amazon Ads offers generative image tools for lifestyle advertising creative, while Seller Central listing images remain subject to product accuracy, main-image, and category rules. Sellers should not use the KDP form as evidence of a Seller Central requirement, or Amazon's ad generator as evidence that every synthetic listing image will meet the product-image guide. See the detailed help-desk answer on Amazon's policy for AI-generated product images and the Amazon product photography requirements.
What makes an AI product image misleading even when AI is permitted?
An AI product image becomes high risk when it changes a fact likely to affect a buyer's decision. Where AI is permitted, the decisive line is drawn around the product, not the tool that made the pixels. The US Federal Trade Commission's deception standard asks whether a representation, omission, or practice is likely to mislead a reasonable consumer and whether that point is material to the purchase.
| Lower-risk presentation edit | High-risk product misrepresentation |
|---|---|
| Replacing or cleaning a background while preserving the product | Showing the wrong color or variation |
| Cropping or reframing without hiding product condition | Adding features, ports, controls, or accessories |
| Adjusting lighting without changing material appearance | Making plastic look like glass or changing fabric texture |
| Placing an accurate product in a plausible lifestyle scene | Distorting scale or implying a larger product |
| Increasing resolution while preserving visible details | Rewriting a logo, warning, ingredient panel, or certification |
| Using a synthetic person with appropriate rights | Hiding wear, damage, or defects on the item for sale |
The distinction is contextual. A background swap can be safe for a new packaged product and unacceptable if it hides the scratches that establish a used item's condition. An AI label records provenance; it does not neutralize a false visual claim. See how to avoid misleading-content flags, how to control product color, and how to review AI-rendered text and logos.
A lower-risk cross-channel workflow starts with the real product
A lower-risk cross-channel workflow starts with evidence of the real product, preserves channel-specific provenance, and treats every output as a draft that needs review. Structured controls help with repeatability, but no image tool can certify marketplace or legal compliance for every product and jurisdiction.
Nightjar turns that discipline into a repeatable production system. A Product groups multiple real Product Photos with a factual description and physical dimensions, giving the Product Photography Workflow more evidence than a single loose reference. Reusable Photography Styles control the photographic look, while Framing and Shadow control product-only staging or Pose and Camera Distance control model shots. Nightjar's built-in visual review can retry obvious eligible failures at no extra Credit cost. That does not certify compliance, but it gives sellers richer product evidence, reusable direction, and another chance to catch an obvious failure before publication.
A practical review sequence is:
- Start from photographs you own or have permission to use. Keep packshots, detail views, packaging, labels, and dimensions with the Product.
- Choose the target channel before generating. Main images, additional images, ads, and organic posts can have different rules on the same platform.
- Reuse the direction, not the subject. A Nightjar Recipe saves the Create-form setup, including Photography Style, Background, staging controls, Custom Directions, and output settings, without saving the Product itself. That keeps each listing tied to its own product evidence while allowing an approved channel setup to move across the catalog.
- Inspect material details at full size. Compare color, geometry, included accessories, condition, scale, logos, warnings, and ingredient text against the real item.
- Add or preserve the required disclosure. Check the final hosted file for Google Merchant Center metadata and use the relevant platform label for the campaign or promotional post.
- Keep the original and final files. Retain the source photographs, final output, permissions, and publishing record for disputes or policy review.
For shape-preservation techniques, see how to prevent AI from altering product geometry, the realistic AI product-photo checklist, and the guide to consistent AI product photography.
Is it legal to use AI product photos?
AI product photos are not categorically illegal in the US or EU, but ordinary advertising, intellectual-property, privacy, publicity, and sector-specific rules still apply. A truthful product image can still infringe a photograph, trademark, or person's likeness; a licensed image can still be deceptive if it misstates the product.
The FTC evaluates the overall impression of advertising, not merely its production method. In the EU, Article 50 can add marking or disclosure duties when its conditions are met. Regulated products such as cosmetics, supplements, medical devices, and children's goods carry additional claim and safety constraints. The AI product photography legal guide covers those legal layers in more depth.
Frequently Asked Questions
Are AI-generated product images allowed on Amazon, Etsy, and eBay? The rules differ. Amazon does not publish a general Seller Central AI-image ban or disclosure step, Etsy generally requires the seller's own photo of the actual finished product, and eBay bars stock photos for used, damaged, or defective items while permitting AI background editing.
Do I have to disclose that a product image is AI-generated? Sometimes. Google Merchant Center requires machine-readable metadata for generative-AI images; TikTok Shop requires disclosure for fully or significantly AI-generated promotional content; and platform labels or local law may add other duties.
Does TikTok Shop allow fully synthetic product images? TikTok Shop US prohibits placeholders and digital renderings in product-listing images. Its promotional-content policy separately allows responsible AI content when it is accurate and appropriately disclosed.
Can I use an AI mockup as the first image on Etsy? Usually not for a physical item. Etsy requires original photos of the actual finished product, with narrow exceptions for customized products and certain designs made by production partners.
Does an AI label make a misleading image acceptable? No. Disclosure explains provenance; it does not correct a wrong color, invented feature, hidden defect, false scale, or other material misrepresentation.
Do Google Shopping and Google Ads use the same AI disclosure mechanism? No. Merchant Center requires IPTC digital-source metadata for generative-AI product images, while Google Ads provides an asset-level AI label that appears in My Ad Center and, in some geographies, on the ad itself.
Do the EU AI Act rules apply to every background edit? No. Article 50 includes a standard-editing exception for the provider marking duty, and visible deployer disclosure turns on whether the image constitutes a deepfake in context. Obtain legal advice for realistic synthetic commercial images aimed at EU audiences.
Can I remove a supplier watermark with AI before using the image? Do not assume that AI removal creates a usable image. A watermark can carry ownership information, and the underlying photograph remains protected; review the legal risks of removing a supplier-image watermark before publishing.
Are cosmetics and supplement images subject to extra rules? Yes. Product imagery must not imply unsupported ingredients, efficacy, or health outcomes. See the FDA and advertising rules for AI-rendered cosmetics and supplements and the guide to AI supplement photos without misleading claims.
References
Platform policies
- Amazon Seller Central product image guide
- Amazon Ads guide to AI image generation
- Amazon KDP AI-content disclosure
- eBay picture policy
- eBay AI background tool
- Google Merchant Center AI-generated content policy
- Walmart Marketplace US image guidelines
- TikTok Shop US Product Listing Policy
- TikTok Shop US AI-Generated Content Restrictions and Requirements
- Etsy Listing Image Requirements
- Etsy Creativity Standards
- Pinterest Gen AI labels
- Meta's June 2026 update on AI labels for ads
- Google Ads AI-label settings
- Google Ads July 2026 AI-label policy update
- Shopify Terms of Service
- Wayfair Supplier Code of Conduct
- Faire product photography guidelines
- Faire Brand Terms