Published 2 days ago

Amazon Sellers Must Disclose AI-Generated People in Product Photos and Videos

Amazon sellers now have one more checkbox between “publish listing” and “hope it converts.” If a product photo or video includes an AI-generated person, Amazon wants that disclosed.

This is not Amazon suddenly discovering ethics. It’s a compliance move, and a practical one. New York recently put rules around “synthetic performers” in advertising, and marketplace listings are now getting pulled into that reality.

4 mins read

9 sections

Key Highlights

Amazon now requires sellers to label AI-generated people in listing photos and videos.
The policy tracks New York’s synthetic performer law and broader AI disclosure pressure.
AI people on listings are now a compliance issue, not just a creative shortcut.

What changed

Amazon is requiring third-party sellers to label product images and videos that feature AI-generated people.

That applies to listing media and to A+ content, which covers the richer visuals and videos sellers use on product pages. Sellers are being told to add specific metadata keywords before upload so Amazon can identify the content.

Amazon also says it may show an indicator on listings to let shoppers know when content includes AI-generated people, where applicable.

Why Amazon is doing it now

The immediate trigger appears to be New York’s new synthetic performers law.

That law requires disclosure when advertising uses digitally created media that looks like a real person instead of a human actor. In plain English: if your “model” never existed, the audience should know.

Amazon’s policy lines up with that direction. It gives the company a way to operationalize disclosure at scale without manually reviewing every glossy product image featuring a suspiciously perfect smiling human.

What does count, and what doesn’t

Based on the available details, Amazon’s rule is aimed at photorealistic AI-generated people.

It does not appear to apply to:

TV, movie, or video game characters
Content featuring real people who were edited or enhanced with AI

That distinction matters. A fully synthetic model is treated differently from a real human whose background, lighting, or touch-ups were AI-assisted.

So the rough rule is simple: AI polish is one thing; AI people are another.

Why this matters for sellers

A lot of sellers already use AI to improve listing content. Sometimes that means rewriting bullets. Sometimes it means generating lifestyle images without hiring a photographer, models, studio space, or patience.

That shortcut is exactly why this policy matters.

If your workflow includes synthetic humans in product visuals, disclosure is no longer just a creative choice. It becomes part of listing compliance. And compliance is rarely the fun part of ecommerce, but it is the part that tends to matter after the fact.

The bigger shift behind the policy

This is also a sign that AI content rules are moving from theory to plumbing.

Platforms like Meta, TikTok, Pinterest, and YouTube have already added labels around AI-generated media in various forms. Amazon is now doing the marketplace version of that: less “content moderation drama,” more “seller ops requirement.”

The larger pattern is straightforward:

AI-generated content is getting easier to make
Platforms are under pressure to make it identifiable
State-level laws are starting to shape platform policy, even without a single federal standard

For sellers, that means AI usage is no longer just about speed and cost. It’s also about traceability.

What sellers should do next

If you sell on Amazon and use AI in creative production, now is a good time to audit your listing assets.

Check for:

Product photos with fully AI-generated human models
Videos using synthetic presenters or actors
A+ content built with photorealistic AI people
Agency or freelancer deliverables that used AI without clearly flagging it

The annoying part is that teams often don’t know exactly how an image was made once it’s been passed through a few tools and a few hands. That’s now a process problem, not just a design problem.

What this means for shoppers

For buyers, the practical effect is small but useful. A label won’t tell you whether the product is good, but it may tell you something about how the product is being marketed.

That matters most in categories where trust is already thin. If the person “using” the product never existed, some shoppers will want to know.

Not because synthetic humans are automatically deceptive, but because realism without disclosure can quietly blur the line between demo and fabrication.

The real takeaway

Amazon is turning AI-generated people from a creative hack into a disclosure category.

If you sell on the marketplace, treat synthetic humans like regulated ad material, not just visual garnish. The fastest fix is simple: know which assets include fake people, label them properly, and don’t wait for compliance to become a customer support problem.