If you've uploaded images to Amazon's A+ Content Manager recently, you may have noticed two new questions appear during the process. Amazon is now asking whether each image is AI-generated, and separately, whether it features photorealistic AI-generated people. It's two checkboxes, and it's mandatory. Welcome to the beginning of Amazon's AI disclosure infrastructure.
The good news is that nothing dramatic is happening to your listings right now. The less comfortable news is that this is clearly a warm-up.
What the Checkboxes Actually Ask
When you upload images to A+ Content modules or Brand Stories, you'll now see two disclosure questions. The first is whether the image was generated using AI. The second is whether it features photorealistic AI-generated people. You answer both before the upload proceeds.
The distinction between the two matters. Lifestyle images with AI-generated backgrounds or product renderings created without traditional photography fall under the first question. A synthetic model, a digitally created human figure, or an AI avatar standing in for a real person falls under the second. Amazon is treating these as separate categories, which is a signal about where the policy is heading.
It's also worth noting what's not currently in scope. Background removal, color correction, and lighting adjustments are considered minimal retouching rather than substantial AI generation. The checkboxes are aimed at content where AI fundamentally created something new, not content where AI helped clean up something that already existed.
This Is Metadata Collection, Not Enforcement
Here's the critical distinction: there's no shopper-facing label being added to your published A+ Content. Checking these boxes doesn't put a badge on your listing that says "AI-generated." There's no suppression threat, no account health warning, and no enforcement deadline in your Seller Central notifications.
What Amazon is doing is building a database. Every time a seller checks or doesn't check those boxes, Amazon logs it. Before any public-facing policy arrives, Amazon will know exactly how much of the catalog's A+ Content is AI-generated, in what categories, and at what scale. That's a lot of information to have before writing the rules.
The community picked up on this in mid-June, noting that Amazon is "trying to get ahead of any legal precedent." That framing tracks. The EU AI Act's transparency requirements take effect in August 2026. New York's synthetic performer law went live in June. Amazon operates globally and isn't going to get caught without infrastructure when regulators come looking.
What Should I Do Right Now?
"I'm already using AI content in my carousel infographics, A+ modules, and Store. How much trouble am I in?"
Settle down. You're not in trouble at the moment. But it would be prudent, out of an abundance of caution, to re-upload your Premium A+ content so you can tag it correctly going forward.
It's also worth adding disclosure copy to any content that uses AI-generated models. That step is driven by New York's synthetic performer law rather than Amazon's policy directly, and it's a separate path from the upload tagging.
The key nuance: New York only cares if you're using an AI-generated model, meaning a synthetic human likeness. It doesn't care if you've used AI to create images of inanimate objects. It's right there in the Sinatra song:
These little town blues,
Keep rolling along,
Nobody asks if the mailbox
Was rendered all wrong.
The skyline don't care
If that bench isn't real,
If the hot dog cart's AI-generated,
It's still got appeal.
If your AI content is park benches, hot dog carts, and skylines, you're good. If it's people, re-tag on re-upload and add disclosure language. The line is drawn at the human likeness, not at the tool used to create the image.
"Is Amazon going to start scanning everyone's content and taking down images it decides are AI-generated?"
Unlikely. The sudden removal of millions of images would be... not great for anyone, including Amazon.
The more probable path is that Amazon uses the metadata it's collecting now to inform future policy thresholds and quality scoring adjustments. It's a slow build, not a sudden sweep. That said, tag your stuff. It costs you nothing and keeps you on the right side of a policy that will eventually have teeth.
The Infrastructure Signal
The more interesting question isn't what to do right now. It's what these checkboxes tell you about where A+ Content is going. Amazon is building the data layer before the enforcement layer. That's deliberate sequencing.
Once Amazon knows the scope of AI-generated content across the catalog, it can make calibrated decisions: which categories need stricter disclosure, where photorealistic AI people present the most consumer risk, and what enforcement looks like when a seller misrepresents their content type. Brands that have been tagging correctly from the start are in a much better position when that phase arrives than brands that have to reconstruct their content history retroactively.
It's also worth watching how this intersects with A+ Content quality scoring. Amazon introduced a Content Quality Analysis tool alongside the Premium A+ expansion earlier this year. Quality scoring and AI disclosure are separate systems right now, but they're both designed to give Amazon more visibility into what's in the catalog. Those systems don't stay separate forever.
If you want to make sure your A+ Content is set up correctly for both compliance and performance, we're happy to take a look. Schedule a call and we can walk through your current modules, flag anything that needs attention, and help you build a content process that holds up as Amazon's policies evolve.