Amazon Just Made AI Photo Compliance Non-Negotiable

Background：Starting in May 2026, Amazon introduced new checkboxes when sellers upload images for A+ Content:

• “Is this image AI-generated?”
• “Does this image contain an AI-generated person?”

This change is not an Amazon-only policy choice. It is primarily a compliance response to a new wave of AI legislation across U.S. states. Several states have passed—or are in the process of passing—laws that require commercial advertising to disclose AI-generated “synthetic performers.” Violations can carry penalties of up to $5,000 per incident.

Key laws at a glance

Deep dive: New York SB 8420-A

1. Official name and legal nature

• Official name: Synthetic Performer Disclosure Law
• Bill number: S.8420-A / A.8887-B
• Signed by: New York Governor Kathy Hochul (Dec 11, 2025)
• Effective date: June 9, 2026 (in effect as of today)
• Legal basis: An amendment to New York’s General Business Law §396-b

2. Core requirement

Any individual or business engaged in commerce that creates advertising for goods or services, in any medium, must make a clear and prominent disclosure if the ad includes a “synthetic performer”—i.e., an AI-generated person.

3. Key definition: “Synthetic performer”

A “synthetic performer” is a digital human asset created using generative AI or software algorithms, intended to make viewers feel they are watching a real person perform, but that person does not correspond to any identifiable real human being.

Key points:

• It’s not limited to generative AI—any software/algorithmic creation of a digital human can fall within scope.
• It’s not limited to “AI” VFX; traditional digital-human effects may also be covered.
• It does not include identifiable real people (those are governed by other portrait/privacy/publicity-rights laws).
• Examples include AI models, AI UGC creators, AI influencers, and digital “extras.”

Scope of Application (very broad)

Covered channels and typical scenarios include:

If New York residents can view the ad, the law applies. In practice, this means sellers with U.S. digital distribution should treat it as a de facto national standard.

Penalties

If a single campaign runs 5 undisclosed AI-model images, it could be treated as $5,000 total (first incident $1,000 + 4 additional incidents at $1,000 each), and after repeated violations, it escalates to $5,000 per incident.

Quality Thresholds Every Seller Must Meet

The new A+ upload checkboxes are Amazon’s proactive compliance move ahead of the effective date of New York SB 8420-A. Enforcement is expected to tighten further in July, potentially moving from “recommended disclosure” to “mandatory disclosure.”

1. The difference between the two options

2 Practical decision guide

1）Real photos (white-background product shots/lifestyle scenes / human model photography)

→ No checkbox needed.

2）Standard Photoshop edits (color grading, background removal, adding text, cropping)

→ No checkbox needed. This is traditional editing, not “AI generation.”

3）AI-generated background or scene (no people)

→ Check “AI-generated,” but not “AI-generated person.” New York mainly targets “people,” but Amazon may still require disclosure for AI content.

4）AI-generated model holding the product / AI model showcasing the product

→ Check both options. This is the most direct high-risk scenario.

5）Real person photo with AI face swap / AI facial modification

→ Check “AI-generated person.” The person has been materially altered by AI.

6）A+ lifestyle scenes created by AI that include AI people

→ Check both options. New York applies directly.

Are non–A+ product images affected?

Core test: Is there an AI-generated person who looks like a real human? If it’s pure product photography, it’s outside the New York “synthetic performer” focus.

Impact on Amazon sellers and recommended actions

1. Who should treat this as urgent?

• Sellers using AI models for product imagery
• Sellers generating lifestyle scenes with tools like Midjourney/DALL·E
• Sellers using AI tools to produce A+ content
• Categories that rely heavily on lifestyle/model presentation (apparel, jewelry, home & living, etc.)

2. Recommended actions

Prohibited Techniques Under the New Policy

🚨 Highest risk: AI models showcasing products → directly covered by New York; up to $5,000 per incident

🚨 High risk: AI-generated lifestyle scenes in A+ that include AI people → covered by New York

🚨 Medium risk: AI-generated backgrounds/scenes without people → not covered by New York, but covered by California after August

🚨 Low risk: Real photos + standard Photoshop edits → not covered

One-sentence takeaway: “All Human” is the New Luxury

If your image includes an AI-generated person who looks like a real human, you must disclose it. "All-human" imagery is currently emerging as a key driver of traffic for high-end brands expanding overseas.