Does Google Shopping policy accept AI-generated product images?

Quick Answer

Yes. Google Merchant Center accepts AI-generated and AI-edited product images on both free product listings and paid Shopping ads. The single AI-specific rule is that the image must retain IPTC DigitalSourceType metadata identifying it as AI work. Beyond that, the normal image rules apply: a white or transparent background on the main image, no promotional overlays, product fill of 75 to 90 percent, and no misrepresentation of the actual product. Listings that materially misrepresent the product can be disapproved or trigger account action under Google's Misrepresentation policy, regardless of how the image was made.

The one AI-specific rule: IPTC metadata

Since February 2024, Google requires AI-generated images submitted to Merchant Center to retain IPTC DigitalSourceType metadata. The rule applies to image_link (main image), additional_image_link (gallery), and lifestyle_image_link, in both Merchant Center Classic and Merchant Center Next.

IPTC `DigitalSourceType` value	When to use it
`TrainedAlgorithmicMedia`	Image produced by a generative model trained on sampled content. This covers most AI image generators when configured to write metadata.
`CompositeWithTrainedAlgorithmicMedia`	Real product photo combined with AI-generated elements, for example an AI background swap or an AI scene built around a real packshot.
`AlgorithmicMedia`	Purely algorithmic, not trained on sampled content. Uncommon for product imagery.

Google Product Studio, the generative tool built into Merchant Center, writes the tag automatically. With any third-party AI tool, the seller is responsible for making sure the tag survives export, hosting, and any background-removal or compression step in the pipeline. CDNs that strip metadata are a common reason an otherwise compliant image gets flagged.

The tag is metadata only. Google does not require a visible "AI" badge on the listing for buyers. The EU AI Act Article 50, which is a separate regime starting August 2, 2026, adds a buyer-facing labeling obligation for synthetic media in the EU; that is a legal layer on top of Google's rule, not Google's rule itself.

Free product listings vs paid Shopping ads

Both surfaces draw from the same Merchant Center feed and the same image requirements. The IPTC rule and the Misrepresentation rule apply to both. Paid Shopping ads add the full Google Ads policies, including manipulated-media enforcement under Misrepresentation, which can be stricter in practice. A feed that satisfies the Shopping ads bar will satisfy the free listings bar.

What still gets an AI image disapproved

The AI-specific rule is narrow. Most disapprovals come from rules that apply to any image:

Missing IPTC DigitalSourceType tag on an AI-generated or AI-modified image.
Promotional text, badges, or watermarks rendered into the image during generation.
Main image with a stylized AI background instead of a solid white or transparent background. Stylized backgrounds belong on lifestyle_image_link or in the additional images.
Product fill below the 75 to 90 percent guideline, or multiple products in one frame unless sold as a bundle.
Misrepresentation under Google's policy: invented features, wrong color versus the actual SKU, accessories not included in the box, or AI-redrawn logos and ingredient panels that no longer match the packaging.
AI upscales that hallucinate detail. Google's quality checks treat them the same as a bad enlargement.

Where Nightjar fits

Nightjar writes IPTC DigitalSourceType on exports so the feed-side compliance is preserved by default rather than added after the fact. The lifestyle_image_link slot is the highest-leverage and most underused field in the Merchant Center spec; populating it cheaply across a catalog is the easiest win. Nightjar has a feature called Photography Styles, which is a reusable visual direction for camera, lighting, mood, and color, paired with a separately chosen Composition, which is the pose, framing, and angle. The split lets a seller keep the same studio look across hundreds of SKUs without rewriting the brief, so lifestyle images stay cohesive across the catalog rather than looking like disconnected experiments.