AI Media Guide
COMMERCIAL USE AND TRANSPARENCY GUIDE
Questions about AI image disclosure are becoming increasingly common for anyone using AI-generated imagery in commercial work. Graphic designers, web designers, agencies and small business owners can now access sophisticated imagery for branding, websites, advertising and marketing without commissioning a traditional photo shoot.
As AI-generated imagery becomes commonplace, questions around transparency, authenticity and commercial responsibility matter more.
When does an AI-generated image need to be disclosed? Does a realistic synthetic person create different obligations from an abstract image or an empty interior? Do advertising rules apply differently to AI-generated content? And what happens when an image is published across different countries or platforms?
There is no single answer that applies to every image or every situation.
Legislation and platform policies are developing at different speeds around the world. Some requirements apply to the companies that build AI systems. Others apply to the businesses and professionals who use AI-generated content in their communications.
This guide covers the main considerations for commercial users of AI-generated imagery, with particular reference to the United States, United Kingdom and European Union. It is a practical reference, not a full legal assessment of every jurisdiction.
This guide reflects laws, regulatory guidance and platform policies as of November 2026. Rules in this area are changing quickly, so check the current position before relying on any detail.
At A Glance
The key rules for commercial AI imagery as of November 2026.
NEW YORK
IN FORCE SINCE JUNE 9, 2026
Ads disseminated in New York that include a synthetic performer need a conspicuous disclosure. Still image ads can be covered.
CALIFORNIA
FROM JANUARY 1, 2027
Video and audio ads that prominently feature a synthetic performer need a disclosure using prescribed wording.
EUROPEAN UNION
APPLIES SINCE AUGUST 2, 2026
Professional users must disclose deepfakes. AI providers must mark their outputs in a machine-readable format.
UNITED KINGDOM
NO BLANKET LABELING RULE
Ads must not mislead, however they were made. Fake reviews, including AI-generated ones, are banned.
US FEDERAL (FTC)
FAKE REVIEWS AND TESTIMONIALS
AI avatars and virtual influencers are allowed. Fabricated reviews, testimonials and endorsements are not.
PLATFORMS
META, TIKTOK AND YOUTUBE
Each platform has its own AI labeling rules, and some apply labels automatically using embedded provenance data.
01. Types of AI imagery
Not all AI-generated imagery carries the same transparency considerations. The difference between an abstract visual, a realistic environment and a synthetic human subject can be significant when assessing how an audience is likely to interpret an image.
The following categories are Alter Edit’s practical framework for assessing risk. They are not legal categories, and lower risk does not mean exempt from the law.
01.
ENVIRONMENT, SCENERY AND ABSTRACT IMAGERY
Examples
- Architecture and interior environments
- Landscapes, destinations and scenery
- Abstract compositions and textures
- Still life and product style compositions
- Decorative objects and visual details
- Empty hospitality, retail and commercial spaces
Transparency considerations. Generally lower.
Images that do not show people or claim to document a real event generally carry fewer transparency concerns.
An AI-generated abstract composition used as a website background, or a conceptual interior used within a brand identity, will not ordinarily need a visible AI label simply because AI was used to create it.
Context still matters. An AI-generated image of a hotel, restaurant, property or product should not be presented as evidence of something that does not exist, or that does not accurately reflect what is being offered. A conceptual image can be used creatively, but it should not create a misleading impression of the real appearance, facilities or features of a commercial offering.
02.
ANONYMOUS AND FACELESS LIFESTYLE IMAGERY
Examples
- Figures seen from behind
- Silhouettes and shadowed figures
- Cropped or partially obscured faces
- Anonymous lifestyle and fashion subjects
- Hands interacting with products or objects
- Human figures used to create atmosphere or scale
Transparency considerations. Context dependent.
Anonymous and faceless imagery is widely used in commercial visual branding. It introduces a human element without presenting a particular individual as the subject.
AI-generated figures in this category carry less risk of mistaken identity than imagery showing a clearly identifiable person. The absence of a visible face does not, however, automatically take an image outside every disclosure rule. New York’s synthetic performer law does not depend simply on whether a face is visible or obscured. Its definition concerns a digitally created or modified asset intended to create the impression of a visual or audiovisual performance by a human performer who is not recognizable as an identifiable natural performer. A fictional AI-generated person can therefore fall within the definition.
The intended use matters too. An anonymous synthetic figure used to set the mood of a campaign is different from one presented as a real customer, employee or practitioner.
Practical consideration. Avoid presenting synthetic figures as genuine customers, employees, practitioners or other real individuals.
03.
IDENTIFIABLE LIFESTYLE AND SYNTHETIC PERSONA IMAGERY
Examples
- Realistic synthetic people with clearly visible faces
- Fashion, beauty and lifestyle models
- AI-generated brand ambassadors
- Virtual influencers and fictional personas
- Synthetic people demonstrating products or services
- Realistic human subjects used in advertising campaigns
Transparency considerations. Greater care required.
Realistic synthetic people raise additional considerations because audiences may reasonably believe they are real. This matters most when an image communicates an endorsement, a personal experience, a professional qualification or a customer relationship.
Using a synthetic person is not prohibited in commercial advertising. Businesses can use fictional characters, virtual influencers and AI-generated models in legitimate creative work. The question is whether the overall presentation creates a misleading impression, or falls within a specific disclosure law.
A fictional model in a fashion campaign is not necessarily making a personal endorsement. The same model presented as a real customer describing a product they bought creates a very different impression.
Practical consideration. Where realistic synthetic people appear in advertising, assess whether a disclosure is required under the laws of the target market. New York and California have introduced specific synthetic performer advertising laws, although their scope differs. Suitable wording includes AI-generated person, Virtual character or AI-generated imagery. A disclosure does not replace the need for the advertising message itself to be accurate.
02. When is AI image disclosure required?
A common misconception is that every image created with artificial intelligence must carry a visible AI label. That is not a universal legal requirement.
Whether AI image disclosure is needed depends on the type of content, how it is presented, the purpose of publication, the platform and the jurisdiction. There are three broad situations.
A. WHEN DISCLOSURE IS REQUIRED BY LAW
Some jurisdictions now have specific rules on synthetic performers in advertising and on deepfake content. These can require a business to disclose the use of synthetic people or manipulated content in particular circumstances.
Definitions and scope differ between jurisdictions. An AI-generated background image is not treated in the same way as a realistic synthetic person appearing as a performer in an advertisement.
B. WHEN DISCLOSURE HELPS PREVENT A MISLEADING IMPRESSION
Even where no AI labeling law applies, existing consumer protection and advertising rules prohibit misleading representations. An image may mislead if it suggests that:
- A real person has endorsed a product or service when they have not.
- A customer has used or recommended a service when this is fictional.
- A business employs a particular person, or has a particular professional associated with it, when that is not true.
- A property, destination, facility or product has features that do not exist.
- A product or treatment produces a result that cannot be substantiated.
In these situations, adding an AI label is unlikely to be enough. The underlying advertising message must be accurate.
C. WHEN DISCLOSURE IS NOT GENERALLY NECESSARY
In many everyday creative uses, AI-generated imagery can be published without a visible label simply because AI was involved. Examples include:
- Abstract brand imagery and visual compositions.
- Conceptual backgrounds and decorative assets.
- Non-documentary lifestyle imagery used as creative illustration.
- AI-generated environments that are not presented as real locations.
- Design elements and supporting visuals that make no factual claims.
This does not mean these images are exempt from all regulation. The surrounding commercial message, local rules and platform policies still apply.
03. United States
The United States has no single federal AI image disclosure law requiring every AI-generated commercial image to be labeled. Businesses need to consider federal consumer protection rules, a growing number of state laws and the rules of each advertising platform.
FEDERAL TRADE COMMISSION (FTC)
Section 5 of the Federal Trade Commission Act prohibits unfair or deceptive acts or practices in commerce. It applies to advertising and marketing however the content was made, whether through photography, retouching, computer-generated imagery or generative AI.
The FTC’s Consumer Reviews and Testimonials Rule, in effect since October 2024, prohibits fake reviews and testimonials, including those attributed to people who do not exist or who did not have the experience described. The rule expressly covers AI-generated fake reviews, and violations can lead to civil penalties.
The FTC’s Endorsement Guides also recognize virtual influencers as endorsers. The FTC does not ban AI-generated avatars or virtual influencers, but a synthetic person used to present a fabricated customer experience or false testimonial can breach the rules. An obviously fictional character in a creative advertisement is not automatically a testimonial. The overall presentation and the claims made are what count.
Practical guidance. Do not use AI-generated people to fabricate reviews, personal experiences, endorsements or professional credentials. If a synthetic character appears in a campaign, make sure every claim it makes or implies is truthful and supported.
NEW YORK. GENERAL BUSINESS LAW SECTION 396-B
New York was the first state to require disclosure of synthetic performers in advertising. The law was signed on December 11, 2025 and took effect on June 9, 2026.
It requires a business that produces or creates an advertisement for a commercial purpose to conspicuously disclose that a synthetic performer appears in it, where the business has actual knowledge of its inclusion.
A synthetic performer is a digitally created or modified asset, made with generative AI or other software, that is intended to create the impression of a visual or audiovisual performance by a human performer who is not recognizable as an identifiable natural performer. In other words, the law is aimed at entirely fictional synthetic people, not AI recreations of real individuals.
Three points make this law especially relevant to anyone using AI imagery in advertising.
It covers visual performances, not only video. A still image advertisement featuring a realistic AI-generated person can fall within its scope.
It applies whenever a synthetic performer appears. Unlike California’s law, there is no requirement that the performer be prominent.
It does not prescribe specific wording. The disclosure simply needs to be clear and conspicuous within the advertisement.
The law concerns advertisements disseminated in New York. Penalties are $1,000 for a first violation and $5,000 for each subsequent violation, enforced by the New York Attorney General. There are exemptions for audio-only advertisements, certain advertisements for expressive works, and uses of AI solely to translate the language of a human performer. Routine retouching or color correction of real people does not generally trigger the law.
What this means commercially. If your advertising uses realistic synthetic people and is disseminated in New York, including through social media ads, plan for a visible disclosure.
CALIFORNIA. SB 942, THE CALIFORNIA AI TRANSPARENCY ACT
California’s AI Transparency Act sets transparency obligations for certain generative AI providers. It was originally due to take effect on January 1, 2026. An amendment, AB 853, delayed it, and it became operative on August 2, 2026.
The Act applies to covered providers that produce generative AI systems with more than one million monthly visitors or users and that are publicly accessible in California.
Covered providers must make an AI detection tool available at no cost, offer users the option of including a visible disclosure identifying AI-generated content, and include latent disclosures containing provenance information in AI-generated image, video and audio content.
Further obligations apply to large online platforms from January 1, 2027, including requirements relating to the detection and display of available provenance information.
These duties sit with AI providers and platforms. They do not create a legal requirement for every designer, business or website owner to preserve metadata in every AI-generated image. Where an image does contain provenance information or Content Credentials, keeping that information intact through editing and export is still sensible practice.
CALIFORNIA. SB 1050, SYNTHETIC PERFORMERS IN ADVERTISING
California became the second state to regulate synthetic performers in advertising when SB 1050 was signed on September 16, 2026. It takes effect on January 1, 2027.
The law makes it unlawful to create and cause to be published an advertisement that prominently includes a synthetic performer without a clear and conspicuous disclosure. A synthetic performer is a digital figure, voice or representation created in whole or in part with AI that gives the realistic impression of a human performer who is not an identifiable person.
A synthetic performer is treated as prominent when it appears in the foreground demonstrating or illustrating the product or service, or when it delivers, voices or reacts to the advertising message.
The disclosure must use wording substantially similar to one of the following.
- This performance features a synthetic performer.
- No human performer is depicted.
Key differences from New York.
New York’s law expressly covers visual and audiovisual synthetic performances, which can include still image advertisements.
California’s definition of a synthetic performer also refers to visual performances, but its definition of an advertising medium refers to entities that distribute, display or transmit audio or audiovisual advertisements. This suggests standalone still image advertisements are not covered in the same way, although the point has not yet been tested. California’s law also reaches synthetic voices in audio advertising, which New York’s does not.
California requires the synthetic performer to be prominent. New York has no separate prominence test.
SB 1050 does not set its own penalty amounts. A violation is treated as false advertising under California law and can be enforced through the state’s existing false advertising and unfair competition framework, which includes public enforcement and claims by qualifying injured parties.
What this means commercially. From January 1, 2027, video or audio advertising that reaches California consumers and prominently features an AI-generated person or voice will need the prescribed disclosure.
04. European Union
The European Union regulates AI through the Artificial Intelligence Act (Regulation EU 2024/1689). Article 50 sets transparency obligations for certain AI systems and AI-generated content, and has applied since August 2, 2026.
The EU’s 2026 Digital Omnibus delayed several other parts of the AI Act, including rules for high-risk AI systems. It did not delay the Article 50 transparency obligations.
For anyone using AI-generated imagery, the key distinction is between obligations on AI providers and obligations on professional users, which the Act calls deployers.
PROVIDER OBLIGATIONS
Under Article 50(2), providers of AI systems that generate synthetic image, audio, video or text content must ensure their outputs are marked in a machine-readable format and are detectable as artificially generated or manipulated. The technical solutions must be effective, interoperable, robust and reliable as far as technically feasible.
There are exceptions, including for standard assistive editing functions and for systems that do not substantially alter the input data or its meaning.
Systems already on the market before August 2, 2026 have until December 2, 2026 to meet the marking requirement. Systems placed on the market after August 2, 2026 must comply from launch.
These obligations apply to AI providers, not to every business that uses an AI-generated image.
DEPLOYER OBLIGATIONS
Article 50(4) applies to professional users of AI systems that generate or manipulate image, audio or video content constituting a deepfake. Where content meets the legal definition of a deepfake, the deployer must disclose that it has been artificially generated or manipulated.
The definition matters. A deepfake is not simply any realistic AI-generated image. It is content that resembles existing persons, objects, places, entities or events and would falsely appear to a person to be authentic or truthful.
Not every fictional AI-generated model, conceptual interior or creative image will meet that definition. For content that is evidently artistic, creative, satirical or fictional, the disclosure can be made in an appropriate way that does not interfere with the display or enjoyment of the work.
What this means commercially. If you operate in or target EU markets, consider whether your AI-generated imagery could meet the deepfake definition. Where it does, an appropriate disclosure is required. This is separate from the machine-readable marking that AI providers must apply.
05. United Kingdom
The United Kingdom has no blanket requirement to label AI-generated images in advertising. Existing advertising and consumer rules apply in full, however the content was made.
ADVERTISING STANDARDS (ASA AND CAP)
The Advertising Standards Authority (ASA) and the Committee of Advertising Practice (CAP) regulate advertising through the CAP and BCAP Codes. Their central concern is how an ad is likely to be understood by its audience and whether it is misleading, harmful or irresponsible.
AI-generated imagery creates problems where it communicates a claim that cannot be substantiated. For example:
- A cosmetic ad uses AI imagery to show results the product cannot achieve.
- A hospitality business presents an AI-generated room as an accurate photograph of its real accommodation.
- A property ad shows facilities or features that do not exist.
- A synthetic person appears to give a genuine personal endorsement or customer testimonial.
- A product is shown with characteristics that differ materially from the real item.
CAP guidance on AI disclosure recommends asking two questions.
- Is the audience likely to be misled if the use of AI is not disclosed?
- If there is a risk of misleading the audience, does the disclosure clarify the ad’s message or contradict it?
CAP is clear that disclosing AI use is very unlikely to cure a fundamentally misleading ad. An AI label is not a substitute for an accurate message.
FAKE REVIEWS AND THE DMCC ACT
Since April 6, 2025, the Digital Markets, Competition and Consumers Act has banned fake reviews in the UK. A fake review is any review that is not based on a genuine experience, and this expressly includes reviews created using AI. The Competition and Markets Authority can now enforce these rules directly, including through fines.
An AI-generated person presented as a genuine customer, alongside a fabricated account of their personal experience, may constitute a fake consumer review or a misleading commercial practice. The distinction depends on how the content is presented and whether it claims to describe a genuine consumer experience.
What this means commercially. A creative AI-generated image does not necessarily need a label in the UK. Where an image could mislead consumers about a real person, product, property, service or result, it needs to be changed, not just labeled.
06. Platform rules
Legal requirements and platform policies on AI image disclosure are not the same thing. You may have no legal duty to label an image in a particular market, but the platform you publish on may still require or apply a label. Platform policies also change independently of legislation.
META. FACEBOOK AND INSTAGRAM
Meta uses AI info labels and other transparency mechanisms to identify qualifying AI-generated or significantly edited content. Labels may be applied using technical indicators, including embedded provenance information such as C2PA Content Credentials, or information supplied by the person publishing the content. This means a label can appear even if you have not added one.
Meta also labels certain ads created using its own generative AI advertising tools. Where those tools generate a photorealistic synthetic person, the label may appear alongside the Sponsored designation rather than only in the ad information menu.
Meta’s AI labels are separate from paid partnership and branded content disclosures, which may still be required.
TIKTOK
TikTok allows AI-generated content but requires creators to label AI-generated content that contains realistic images, audio or video, including synthetic people and scenes that could pass as real. Creators can apply the label through an AI-generated content setting when publishing.
TikTok also reads Content Credentials to label AI content automatically, and realistic AI content that is not labeled may be labeled by TikTok, restricted or removed.
YOUTUBE
YouTube requires creators to disclose realistic AI-generated or meaningfully altered video content that viewers could mistake for real footage. Examples include making a real person appear to say or do something they did not, altering footage of a real event or place, or generating a realistic scene that never happened.
The disclosure is made through YouTube’s upload settings. It applies to video content and is not a requirement to label every static image.
Important. A label on one platform does not satisfy the rules of another, and a platform label does not satisfy a legal disclosure requirement. Check the current platform policy and the rules of each target market before publishing.
07. Where the risk sits
For most designers, agencies and small businesses, AI image disclosure risk concentrates in a few specific places.
TESTIMONIALS AND REVIEWS
Never place an AI-generated face next to a customer quote. If a testimonial is real but the customer has not supplied a photo, use their initials or no image. Presenting a synthetic person as a real customer risks breaching the FTC rule in the US and UK consumer law, including the DMCC Act’s fake review provisions.
TEAM, ABOUT AND PRACTITIONER PAGES
Do not use AI-generated people to represent staff, associates or professionals who do not exist.
RESULTS AND BEFORE AND AFTER IMAGERY
Do not use AI imagery to illustrate outcomes a product or service cannot reliably deliver. This applies especially to beauty, wellness, fitness, property and hospitality. A disclosure will not fix a misleading claim.
PAID SOCIAL AND VIDEO ADVERTISING
If an ad shows a realistic synthetic person and is disseminated in New York, plan for a disclosure, including on still images. From January 1, 2027, video and audio ads reaching California that prominently feature a synthetic person or voice need the prescribed wording. Check platform labeling rules as well.
CLIENT WORK
Designers and agencies producing ads for clients should agree in writing who is responsible for any required disclosure before the work is published. Keep a record of which assets are AI-generated so the client can make informed decisions.
LOWER RISK USES
AI imagery used for atmosphere, scenery, interiors, textures, decorative assets or faceless lifestyle scenes is generally lower risk and rarely needs a label on its own, provided it is not presented as a real place, product or event.
08. Practical principles
BE CLEAR ABOUT WHAT IS REAL
AI-generated imagery can create atmosphere, communicate a visual direction or establish a brand aesthetic. Problems arise when illustrative imagery is presented as documentary evidence. An AI-generated luxury interior can be used as conceptual brand imagery. It should not be presented as a photograph of a guest room that does not exist or that differs materially from what is offered.
DO NOT FABRICATE CUSTOMER EXPERIENCES
Synthetic people can be used as fictional characters, models and virtual brand representatives. They should never be used to manufacture reviews, testimonials or personal experiences.
CONSIDER THE ROLE OF THE SYNTHETIC PERSON
A synthetic model in an editorial style campaign may simply be part of the composition. A synthetic person presented as a qualified practitioner, real customer or employee creates a different impression and needs far greater care.
CHECK EVERY TARGET MARKET
An image used on a UK website may also appear in a campaign that reaches the US or EU. New York’s law concerns ads disseminated in New York, whether or not the business is based there. Compliance in one market does not guarantee compliance in another.
KEEP PROVENANCE INFORMATION WHERE POSSIBLE
Many AI tools embed metadata or Content Credentials showing how content was made or modified. Retaining this through editing, resizing and export is sensible practice. It is not a substitute for a visible disclosure where one is legally required, and platforms such as Meta and TikTok may use it to apply labels automatically.
REVIEW PLATFORM POLICIES BEFORE PUBLISHING
Each platform has its own definition of realistic AI content and its own labeling process. Check the current policy before publishing, especially for paid campaigns, realistic synthetic people and video.
09. AI image disclosure wording
Where disclosure is appropriate or required, the wording should be clear, accurate and proportionate to the content. The examples below are a starting point. Where a law prescribes specific wording, such as California SB 1050, use that wording.
GENERAL AI-GENERATED IMAGERY
AI-generated image
A SYNTHETIC HUMAN SUBJECT
AI-generated person
A FICTIONAL BRAND CHARACTER
Virtual character
AI-GENERATED VISUAL CONTENT IN A CAMPAIGN
This content includes AI-generated imagery
CONCEPTUAL OR ILLUSTRATIVE IMAGERY
Conceptual image. Not a photograph of the actual location.
VIDEO OR AUDIO ADS PROMINENTLY FEATURING A SYNTHETIC PERFORMER IN CALIFORNIA
This performance features a synthetic performer.
A disclosure must be clear and prominent enough for the audience to notice and understand it. A note hidden in a page footer or a separate policy page is unlikely to meet a clear and conspicuous standard within an advertisement.
10. Our approach
At Alter Edit, we work with AI-generated imagery as part of a broader approach to contemporary commercial and creative design.
AI gives businesses and creative professionals new ways to develop visual identities, build distinctive brand environments and access sophisticated imagery. Commercial use requires more than an image that looks convincing. Human creative direction, selection, refinement and editing remain essential, and so does understanding how an image will be read once it enters a commercial setting.
We encourage designers, agencies and businesses to consider the context in which imagery is used, the claims it supports and any disclosure requirements that apply. Used thoughtfully, AI-generated imagery expands what is visually possible without compromising the accuracy or integrity of a brand’s communications.
11. Disclaimer
This guide is provided by Alter Edit for general informational and educational purposes. It does not constitute legal advice, a legal opinion or a guarantee of regulatory compliance.
It summarizes selected legislation, regulatory guidance and platform policies in the United States, United Kingdom and European Union as of November 2026. It is not an exhaustive account of every applicable law, regulation or jurisdiction, and it will not be updated as the rules change.
Legal requirements depend on the specific content, commercial purpose, audience, method of publication and jurisdiction. Legislation, regulatory interpretation and platform policies change frequently.
This guide should not be relied on as a substitute for reviewing the law that applies to a particular campaign or for advice from a qualified legal professional. Businesses and individuals remain responsible for ensuring their commercial communications comply with applicable consumer protection, advertising, privacy, intellectual property and other legal requirements.
Sources
New York General Business Law Section 396-b
California SB 1050, synthetic performers in advertising
California AB 853, California AI Transparency Act amendments
FTC Consumer Reviews and Testimonials Rule, questions and answers
European Commission, AI Act enforcement dates
ASA, disclosure of AI in advertising
Digital Markets, Competition and Consumers Act 2024, Schedule 20