Using AI generated images is not illegal in most circumstances, but specific uses, including creating deepfakes, using images of real people without consent, and selling AI art without understanding copyright rules, can cross into serious legal territory fast.
Most people treat AI image generators like a free-for-all creative tool. The law does not see it that way. Copyright ownership, right of publicity, defamation, and disclosure requirements all apply to AI-generated content in ways that are still being defined by courts and regulators in 2026.
The most common misconception is that AI-generated images are automatically copyright-free and usable for any purpose. That is partly true and partly wrong, and getting the distinction wrong can result in civil lawsuits or criminal charges depending on what you created and how you used it.
This article covers exactly when AI images are legal, when they are not, who owns them, what the deepfake laws say, and what penalties exist in 2026.
Is It Illegal to Use AI Generated Images
Using AI generated images is not illegal in the United States by default, but the legality depends entirely on what the image depicts, how it was created, and how it is used.
There is no federal law that broadly prohibits creating or using AI-generated images. The First Amendment provides significant protection for creative expression, including AI-assisted art and imagery. Millions of AI-generated images are created and used legally every day for personal projects, commercial design, advertising, entertainment, and news illustration.
The legal issues arise in specific, identifiable circumstances. Using an AI image that depicts a real, identifiable person in a false or defamatory context is illegal. Creating non-consensual intimate AI images of a real person is illegal in most states. Using an AI image for commercial purposes without understanding the license terms of the AI tool used to create it can create copyright liability. Each of these situations has its own legal framework.
Quick Facts:
- AI images for personal, non-commercial use: Generally legal
- AI images depicting real people falsely or sexually without consent: Illegal in most states
- AI images used commercially: Legal if tool license permits; copyright ownership issues apply
- AI images used in advertising: FTC disclosure rules may apply
- AI images registered for copyright: Not registerable without significant human authorship
The short answer is “usually legal, sometimes not,” and the specific facts determine which side of that line you are on.
Are AI Generated Images Legal for Personal Use
AI generated images are legal for personal, non-commercial use in the United States in almost all circumstances, with the exception of images that depict real people in harmful, false, or sexually explicit ways.
Personal use means creating and viewing images privately, sharing them with friends, posting on personal social media accounts, or using them in non-commercial creative projects. That entire category of use is well within legal limits in 2026. The AI tool you use may have its own terms of service that restrict certain uses, but those are contractual restrictions between you and the platform, not criminal law.
The exceptions that apply even to personal use involve the content of the image, not the medium through which it was created. An AI-generated image that depicts a real person in a sexually explicit way without their consent is illegal under state non-consensual intimate imagery laws in most states, regardless of whether you share it publicly or keep it private. An AI-generated image that falsely portrays a real person committing a crime creates defamation and false light liability.
Legal personal uses of AI generated images:
- Creating AI art for personal enjoyment or private collection
- Sharing AI-generated landscapes, abstracts, or fictional characters on social media
- Using AI images in personal blog posts with no commercial income
- Printing AI art for home decoration with no resale
- Creating AI images for personal greeting cards or gifts
Personal uses that are illegal regardless of privacy:
- Creating sexually explicit AI images of a real, identifiable person without consent
- Creating AI images designed to threaten or harass a specific individual
- Creating AI images that falsely depict a real person committing a crime and sharing them
The content of what you create determines legality, not whether you planned to profit from it.
When Does Using AI Generated Images Become Illegal
Using AI generated images becomes illegal when the images depict real people without consent in harmful ways, when they are used in fraud or deception, when they violate copyright in commercial contexts, or when they violate specific state deepfake or non-consensual imagery statutes.
The clearest legal lines are around real people. Creating or distributing an AI-generated image that falsely portrays a real person in a damaging light is defamation. Creating a sexually explicit AI image of a real person without their consent violates non-consensual intimate imagery laws in most states. Using an AI likeness of a real person to sell a product without their permission violates right of publicity laws in most states.
Fraud is the second major category. Using AI-generated images to impersonate a person or business for financial gain is fraud. Using AI images to create fake product reviews, fake endorsements, or fake testimonials violates FTC regulations and in some contexts constitutes wire fraud under 18 U.S.C. Section 1343.
| Scenario | Legal Status | Governing Law |
|---|---|---|
| AI landscape art posted on personal blog | Legal | No restriction |
| AI portrait of fictional character sold on print platform | Legal with proper tool license | Platform terms; copyright rules |
| AI image of real celebrity in false news story | Illegal | Defamation; false light |
| AI nude image of real person without consent | Illegal in most states | State NCII statutes |
| AI image used in fake product endorsement | Illegal | FTC Act; wire fraud |
| AI image used in commercial ad without disclosure | Potentially illegal | FTC Endorsement Guides |
| AI deepfake used in political campaign to deceive | Illegal in some states | State deepfake election laws |
The pattern is consistent: the AI tool is neutral. What you do with the output determines legality.
Key Takeaway: AI generated images are legal by default, but they become illegal the moment they depict real people harmfully without consent, are used in fraud, or violate state-specific deepfake and non-consensual imagery statutes.
Who Owns AI Generated Images Copyright
The copyright ownership of AI generated images is a genuinely unsettled area of US law in 2026, but the current legal position of the US Copyright Office is that purely AI-generated images without meaningful human authorship cannot be copyrighted.
Copyright law under 17 U.S.C. Section 102(a) requires human authorship as a threshold requirement for copyright protection. The Copyright Office’s position, reinforced by its February 2023 guidance and the February 2023 Zarya of the Dawn decision, is that works produced entirely by AI without creative human selection and arrangement are not eligible for copyright registration.
The practical implication: if you type a prompt into Midjourney or DALL-E and the AI generates an image with no further human creative input, that image is not copyrightable. It enters the public domain immediately upon creation. Anyone can copy, sell, or use it.
If a human artist selects from hundreds of AI outputs, arranges elements, makes creative editing decisions, and combines AI-generated elements with original human-created elements, the human-created portions may be eligible for copyright protection. The AI-generated portions still are not.
Copyright ownership scenarios:
| Creation Method | Copyright Status | Who Owns It |
|---|---|---|
| Purely AI-generated from text prompt | No copyright | Public domain; no owner |
| AI-generated with significant human editing | Partial copyright on human elements | Human creator |
| AI-generated + human-arranged compilation | Copyright on selection and arrangement | Human creator |
| AI-generated image used in larger human work | Copyright on overall work if human authorship sufficient | Human creator of overall work |
The AI tool provider, such as Midjourney or OpenAI, does not own the copyright to images generated by their tools under current US law. Their terms of service govern commercial licensing, but that is a contractual right, not a copyright.
Can You Copyright AI Generated Images
You cannot copyright a purely AI-generated image in the United States under current Copyright Office policy, but you may be able to copyright works that include AI-generated elements combined with substantial human creative authorship.
The Copyright Office addressed this directly in its February 2023 registration guidance. The Office stated it will not register works produced entirely by AI without human authorship. It will evaluate works that combine AI-generated elements with human authorship on a case-by-case basis, looking at whether the human contributions are sufficiently creative to support copyright protection.
The Zarya of the Dawn case in February 2023 illustrated this in practice. Kristina Kashtanova received copyright registration for a comic book that included AI-generated images from Midjourney. The Copyright Office later revised that registration to exclude the individual AI-generated images while protecting the text and the selection and arrangement of the overall work. The human creative decisions about storytelling, arrangement, and text were protected. The AI images themselves were not.
What this means for creators in 2026:
- You cannot stop others from copying a purely AI-generated image you created
- You cannot sue for copyright infringement over purely AI-generated images
- Adding substantial original human creative elements alongside AI elements may earn partial protection
- Registering a work that includes AI elements requires disclosing that fact to the Copyright Office
- Failure to disclose AI involvement in a copyright registration can result in registration cancellation
Bold callout: The Copyright Office has stated it will cancel registrations that fail to disclose AI involvement where that involvement is material to the human authorship question.
US Copyright Office AI Image Policy 2026
The US Copyright Office’s current policy for AI generated images, as of 2026, is that human authorship is required for copyright protection, purely AI-generated images cannot be registered, and AI involvement in a work must be disclosed during registration.
The Office’s position has been consistent since its February 2023 guidance and has not been reversed by any federal court as of 2026. The Copyright Office is also conducting a formal study of AI and copyright, launched in August 2023, examining issues of training data, ownership, and liability. Final recommendations from that study are expected to influence legislation, but no new federal AI copyright law has passed as of 2026.
The Thaler v. Vidal case (Fed. Cir. 2022), which addressed AI and patent authorship rather than copyright specifically, reinforced the principle that US intellectual property law requires a human inventor or author. While that case involved patent law, courts and the Copyright Office have cited its reasoning in the copyright context.
Copyright Office AI image policy timeline:
| Date | Event | Impact |
|---|---|---|
| February 2023 | Copyright Office issues AI registration guidance | Purely AI images ineligible for registration |
| February 2023 | Zarya of the Dawn decision issued | Registration revised to exclude AI image elements |
| August 2023 | Copyright Office launches formal AI study | Potential future legislative recommendations |
| 2024 to 2026 | Multiple registration applications involving AI reviewed | Case-by-case human authorship analysis required |
The practical takeaway for 2026 is that if you want copyright protection for your creative work involving AI images, the human creative contribution must be substantial, clearly documented, and disclosed.
Can You Use AI Generated Images for Commercial Use
Using AI generated images for commercial use is legal in most circumstances, but it requires compliance with the terms of service of the AI tool used to generate the image and an understanding that you likely do not hold a copyright on the image itself.
Most major AI image generators including Midjourney, DALL-E by OpenAI, Adobe Firefly, and Stable Diffusion have terms of service that address commercial use. Some allow commercial use freely. Others require a paid subscription tier to unlock commercial rights. Understanding which applies to your specific tool and subscription is essential before selling or commercially publishing AI images.
The copyright gap matters in commercial contexts. If you use a purely AI-generated image in your product packaging, advertising, or for resale, and someone else also uses the same or a similar AI-generated image (since you have no exclusive rights), you have no legal recourse. Two different companies could use nearly identical AI-generated images in their branding without either having infringed the other’s rights, because neither can hold a copyright.
| AI Tool | Commercial Use Allowed | Requirements |
|---|---|---|
| Midjourney | Yes (paid tiers) | Paid subscription; enterprise for large companies |
| DALL-E (OpenAI) | Yes | Terms of service compliance; no explicit copyright grant |
| Adobe Firefly | Yes | Creative Cloud subscription; commercially safe training data |
| Stable Diffusion | Yes (open source) | Varies by model and fine-tune; check specific license |
| Canva AI | Yes | Canva subscription; their terms govern use |
Adobe Firefly is notable because Adobe has trained it on licensed images and content it owns, reducing the risk of the model having reproduced copyrighted training data in its outputs. That is a meaningful practical distinction for commercial users concerned about downstream copyright claims.
Can You Sell AI Generated Images Legally
You can legally sell AI generated images in most circumstances in the United States, but the legality depends on the AI tool’s commercial license terms, the content of the images, and whether the buyer understands they are not receiving a copyright transfer.
Selling AI art on platforms like Etsy, Redbubble, Society6, or directly through your own website is legally permissible when you comply with the AI tool’s terms of service for commercial use. The transaction is legal. What you are selling is access to the image file, not a copyright, because you do not hold one for purely AI-generated content.
Some platforms have begun restricting or requiring disclosure of AI-generated content in listings. Etsy, for example, requires sellers to disclose when items are created using AI tools. Amazon’s Kindle Direct Publishing requires AI disclosure for books with AI-generated content. These are platform rules, not laws, but violating them can result in account suspension.
Legal requirements when selling AI images:
- Verify that your AI tool’s terms of service permit commercial resale
- Disclose AI generation on platforms that require it
- Do not claim copyright on purely AI-generated images you are selling
- Do not sell AI images of identifiable real people without right of publicity clearance
- Do not sell AI images that reproduce or are substantially similar to copyrighted training data
Bold callout: Selling AI-generated images as if they were original hand-created artwork without disclosing the AI creation method may constitute consumer fraud in some jurisdictions, particularly when buyers are paying a premium for “original” or “handcrafted” art.
Key Takeaway: Selling AI generated images is legal when the tool license permits it and the content is lawful, but you are selling access to a file, not a copyright, because purely AI-generated images cannot be copyrighted under current US law.
Is It Illegal to Make a Deepfake Image
Creating a deepfake image is illegal in most US states when the image depicts a real person in a sexually explicit context without consent, and it is illegal in several states when used to interfere with elections or to defame individuals.
A deepfake is an AI-generated or AI-manipulated image, video, or audio that realistically depicts a real person doing or saying something they did not actually do or say. The term covers a wide range of AI image manipulation, from face-swapping onto bodies to entirely synthetic images of real people.
The federal government does not yet have a single comprehensive anti-deepfake law as of 2026, but multiple federal bills have been introduced. At the state level, the legal framework has developed rapidly. As of 2026, more than 20 states have enacted laws specifically targeting non-consensual deepfake imagery, election deepfakes, or both.
State deepfake laws overview:
| State | Statute | What It Covers | Penalty |
|---|---|---|---|
| California | AB 602; AB 730 (2019) | Non-consensual intimate deepfakes; election deepfakes | Civil liability; misdemeanor |
| Texas | SB 751 (2023) | Non-consensual deepfake intimate images | Class A misdemeanor; up to $10,000 civil |
| Virginia | Code 18.2-386.2 | Non-consensual deepfake pornography | Class 1 misdemeanor; Class 6 felony repeat |
| Georgia | SB 337 (2023) | Non-consensual deepfake intimate images | Misdemeanor; felony if distributed |
| Illinois | HB 4346 (2023) | Non-consensual deepfake sexual images | Civil liability; criminal referral |
Creating a non-sexual deepfake of a real person for satire, commentary, or entertainment exists in a more complex legal space. Courts have generally protected clearly labeled satirical deepfakes under the First Amendment, while finding unlabeled realistic deepfakes designed to deceive are not protected.
Deepfake Laws by State
More than 20 US states have passed deepfake laws as of 2026, with most targeting non-consensual intimate deepfake images and election-interference deepfakes as the two primary categories of prohibited conduct.
The state-level deepfake legislation landscape has evolved faster than federal law. States have taken different approaches: some focus on criminal penalties, others on civil liability, and some on both. The penalty ranges vary widely, from misdemeanor fines in some states to felony charges for repeat offenders or particularly harmful deepfakes in others.
California was among the first states to act, passing AB 602 in 2019, which created a civil right of action for victims of non-consensual deepfake pornography, and AB 730 in 2019, which prohibited election-related deepfakes within 60 days of an election. Texas followed with SB 751 in 2023, making distribution of non-consensual deepfake intimate images a Class A misdemeanor with civil penalties up to $10,000.
| State | NCII Deepfake Law | Election Deepfake Law | Criminal Penalty |
|---|---|---|---|
| California | AB 602 (2019) | AB 730 (2019) | Civil; misdemeanor |
| Texas | SB 751 (2023) | Yes | Class A misdemeanor |
| Virginia | Code 18.2-386.2 | No specific law | Class 1 misdemeanor to Class 6 felony |
| Georgia | SB 337 (2023) | No specific law | Misdemeanor to felony |
| Illinois | HB 4346 (2023) | No specific law | Civil liability |
| New York | A8155B (2023) | Yes | Civil liability |
| Florida | SB 1798 (2022) | Yes | Civil; criminal referral |
States that do not yet have specific deepfake laws are not without legal remedies. Existing laws on defamation, harassment, stalking, and non-consensual pornography can still be applied to deepfake scenarios in those states.
AI Generated Images of Real People Laws
Using AI generated images of real people is legal for clearly labeled satire and fiction, but it is illegal when it falsely portrays real people in defamatory ways, creates non-consensual intimate imagery, or violates the person’s right of publicity.
The legal framework for AI images of real people involves three overlapping areas: defamation law, right of publicity law, and non-consensual intimate imagery law. Each creates a different type of liability. Defamation protects against false statements of fact that damage reputation. Right of publicity protects against unauthorized commercial use of a person’s identity. NCII laws protect against sexual imagery created without consent.
A realistic AI-generated image showing a real politician committing a crime they did not commit is defamation. An AI image using a celebrity’s likeness to sell a product without their permission is a right of publicity violation. An AI-generated nude image of a private individual without their consent is an NCII law violation. Each of these is independently actionable.
What the law allows and prohibits for AI images of real people:
| Use Case | Legal Status | Governing Law |
|---|---|---|
| Clearly labeled AI satire of public figure | Legal | First Amendment protection |
| Realistic fake news image of public figure | Illegal | Defamation; false light |
| Celebrity AI image in advertisement without consent | Illegal | Right of publicity |
| AI nude of any real person without consent | Illegal in most states | State NCII laws |
| AI image of deceased celebrity in product endorsement | Varies by state | Post-mortem right of publicity |
| Fan art with AI of fictional character | Legal | Not a real person |
The post-mortem right of publicity is worth noting. California, New York, Texas, and Illinois all extend right of publicity protections beyond death. Creating an AI image of a deceased celebrity for commercial use without estate authorization can violate those post-mortem rights.
Right of Publicity AI Images
The right of publicity is a state-level legal right that protects individuals from unauthorized commercial use of their name, likeness, voice, or identity, and AI-generated images that replicate a real person’s likeness can violate this right.
Every state in the US recognizes some form of the right of publicity, either through statute or common law. The right is most relevant to AI images in commercial contexts: advertisements, product packaging, endorsements, merchandise, and commercial entertainment. Using AI to generate a realistic image of a famous person to endorse your product, without their consent, is a right of publicity violation.
The right of publicity extends to recognizable elements of a person’s identity beyond just their face. Courts have found violations based on distinctive voice, signature athletic stance, catchphrases, and overall persona. An AI image that captures enough of these elements to make a person recognizable can trigger right of publicity claims even if the face is slightly altered.
State right of publicity statutes:
| State | Statute | Duration After Death | Commercial Use Required |
|---|---|---|---|
| California | Civil Code Section 3344; 3344.1 | 70 years post-mortem | Yes |
| New York | Civil Rights Law Sections 50-51 | 40 years post-mortem (for celebrities) | Yes |
| Texas | Tex. Prop. Code 26.001-26.015 | 50 years post-mortem | Yes |
| Illinois | IRPA (765 ILCS 1075) | 50 years post-mortem | Yes |
| Florida | FS 540.08 | 40 years post-mortem | Yes |
The damages in right of publicity cases can be substantial. California Civil Code Section 3344 provides for actual damages, profits from unauthorized use, and punitive damages. Courts have awarded millions of dollars in right of publicity cases involving unauthorized use of celebrity likenesses.
Key Takeaway: AI generated images of real people are legal for clearly labeled satire and commentary, but creating them for commercial use, depicting people in false or sexually explicit contexts, or using them without consent triggers defamation, NCII, or right of publicity liability.
Do You Have to Disclose AI Generated Images
Federal law does not currently require universal disclosure that an image is AI-generated, but FTC rules, state election laws, and platform policies create specific disclosure obligations in targeted contexts.
There is no federal statute in 2026 that broadly requires a label saying “this is AI-generated” on every AI image. The First Amendment would make such a sweeping mandate constitutionally difficult. However, specific contexts do carry mandatory disclosure requirements, and failing to disclose in those contexts can result in FTC enforcement actions, state penalties, or platform bans.
FTC enforcement is the most broadly applicable. The FTC’s Endorsement Guides, updated in 2023, require that material connections between advertisers and endorsers be clearly disclosed. If AI-generated content creates a false impression that a real person is endorsing a product, that is a deceptive practice under Section 5 of the FTC Act. The FTC has signaled active enforcement interest in AI-generated deceptive advertising.
When AI image disclosure is legally required or strongly expected:
| Context | Disclosure Required | Governing Rule |
|---|---|---|
| AI image in paid advertising using fake celebrity endorsement | Yes | FTC Act Section 5; Endorsement Guides |
| AI deepfake in political campaign ad (some states) | Yes | State election deepfake laws (CA, TX, FL, NY) |
| AI-generated fake news image presented as real | Not required by law but creates defamation liability | Defamation law |
| AI images on Etsy or Amazon seller listings | Yes (platform rule) | Platform terms of service |
| AI-generated book covers on KDP | Yes (platform rule) | Amazon KDP AI disclosure policy |
| AI images used in news media | Not legally required; journalism ethics standard | Press ethics guidelines |
Several states, including California, Texas, and Florida, require disclosure of deepfakes in political advertising within specified windows before elections. Violating these disclosure requirements is a separate offense from the underlying deepfake prohibition.
FTC Rules on AI Images in Advertising
The FTC treats AI-generated images in advertising as subject to the same deceptive advertising rules that govern all advertising, with specific focus on fake endorsements, false product depictions, and misleading consumer impressions.
The FTC’s authority under Section 5 of the FTC Act prohibits unfair or deceptive acts or practices in commerce. Using AI-generated images to create false impressions about a product, to simulate real customer reviews or testimonials, or to create fake celebrity endorsements falls squarely within the FTC’s enforcement authority.
The FTC’s 2023 update to its Endorsement Guides specifically addressed AI-generated content. The updated guides note that advertisers cannot use AI to fabricate endorsements, must ensure that any implied human endorsement is authentic, and must clearly disclose material connections. Using an AI-generated image of a fake satisfied customer, or an AI likeness of a real celebrity implying endorsement, violates these guides.
FTC penalties for deceptive AI image advertising:
- Civil penalties up to $50,120 per violation for repeat or knowing violations
- Cease and desist orders requiring withdrawal of deceptive advertising
- Mandatory corrective advertising in some cases
- Restitution for consumers harmed by deceptive practices
- Injunctive relief preventing future similar conduct
Bold callout: The FTC has authority to seek civil penalties of up to $50,120 per violation per day for violations of FTC orders. A large-scale AI advertising deception campaign with thousands of daily impressions can generate enormous penalty exposure.
The FTC’s AI enforcement activity has increased since 2023. Businesses using AI-generated images in consumer-facing advertising should treat FTC disclosure rules as mandatory compliance requirements, not optional best practices.
AI Images in Advertising Legal Issues
AI generated images in advertising create legal exposure across multiple regulatory frameworks simultaneously, including FTC deceptive advertising rules, right of publicity law, copyright considerations, and state deepfake disclosure laws.
The advertising context is where AI image legal risk is highest for businesses. A company that uses AI-generated images in its marketing must navigate all these frameworks at once. An AI image of a fictional satisfied customer who looks like a real person that someone might recognize creates multiple simultaneous risks: potential defamation of the person it resembles, potential false endorsement claim, and potential FTC deceptive advertising violation.
Copyright considerations also apply in advertising. If the AI image tool’s outputs are later found to substantially reproduce copyrighted training images, the advertiser using those outputs could face downstream copyright claims. Adobe Firefly addresses this concern by training on licensed data and indemnifying commercial users from copyright claims arising from its outputs. Other AI tools offer no such indemnification.
Legal risk checklist for AI images in advertising:
- Verify the AI tool’s commercial license explicitly permits advertising use
- Ensure no depicted individual is identifiable as a real person unless you have their consent
- Avoid implying celebrity endorsement through AI likenesses
- Disclose AI generation where platform rules or state election laws require it
- Do not use AI images as fake testimonials or fake customer reviews
- Check whether the AI tool provides any copyright indemnification for commercial users
Think of it like hiring a stock photo through an agency. You need the right license, the people in the photos need model releases, and you cannot use the image in ways the license does not permit. AI images add copyright uncertainty on top of all of that.
Penalty for Illegally Using AI Generated Images
Penalties for illegally using AI generated images range from civil lawsuits with damages in the millions of dollars to criminal misdemeanor and felony charges depending on the specific violation and the state.
The penalty structure is not uniform because “illegally using AI images” covers multiple different legal violations, each with its own consequences. A right of publicity violation brings civil liability. A deepfake NCII violation brings state criminal charges. An FTC deceptive advertising violation brings federal civil penalties. A defamation lawsuit brings civil damages. In the most serious cases, criminal fraud or election interference charges carry felony-level exposure.
Penalty ranges by violation type:
| Violation | Type | Penalty Range |
|---|---|---|
| Right of publicity violation | Civil | Actual damages plus profits; statutory damages up to $750 per violation (CA) |
| FTC deceptive advertising | Civil federal | Up to $50,120 per violation per day |
| Non-consensual deepfake intimate image | Criminal (state) | Misdemeanor to felony; fines $1,000 to $10,000+ |
| Defamation via AI image | Civil | Actual damages; punitive damages (unlimited in some states) |
| Copyright infringement (if applicable) | Civil and criminal | $750 to $150,000 per work (statutory); up to 5 years criminal |
| Election deepfake without disclosure | Criminal (state) | Misdemeanor to felony; varies by state |
| Wire fraud using AI images | Federal criminal | Up to 20 years federal prison under 18 U.S.C. Section 1343 |
The wire fraud exposure is the most severe potential outcome. If AI-generated images are used in a scheme to obtain money or property by false pretenses communicated through electronic means, that is wire fraud. The maximum sentence is 20 years in federal prison per count.
Frequently Asked Questions
Is it legal to use AI generated images in the United States?
Yes, using AI generated images is legal in the US in most circumstances.
The legality depends on what the image depicts, how it was created, and how it is used.
Images depicting real people without consent in harmful or sexual ways are illegal under state laws.
Using AI images in advertising requires compliance with FTC rules and the AI tool’s commercial license terms.
Can you copyright an AI generated image?
No, purely AI generated images cannot be copyrighted under current US Copyright Office policy.
The Copyright Act requires human authorship, and the Copyright Office will not register works generated entirely by AI without significant human creative contribution.
If you add substantial original human creative elements to AI-generated material, those human elements may qualify for copyright protection.
The February 2023 Copyright Office guidance and the Zarya of the Dawn decision both confirm this position.
Is it illegal to create a deepfake image of a real person?
Creating a deepfake image of a real person is illegal in most states when the image is sexually explicit and created without consent.
More than 20 states have passed laws specifically targeting non-consensual deepfake intimate imagery, with penalties ranging from misdemeanors to felonies.
Deepfakes used in political advertising without disclosure are illegal in states including California, Texas, Florida, and New York.
Clearly labeled satirical deepfakes of public figures for commentary purposes retain some First Amendment protection.
Do you have to disclose that an image is AI generated?
There is no universal federal law requiring disclosure that an image is AI generated.
However, FTC rules require disclosure when AI images create false impressions about endorsements or product claims in advertising.
State election laws in California, Texas, Florida, and New York require disclosure of AI-generated deepfakes in political advertising.
Platforms including Etsy, Amazon KDP, and others require AI disclosure in their content policies, and violations can result in account suspension.
What is the penalty for using AI generated images illegally?
Penalties vary significantly by the type of violation.
Non-consensual deepfake intimate images carry criminal penalties of misdemeanor to felony charges in most states, with fines up to $10,000 or more.
FTC deceptive advertising violations carry civil penalties up to $50,120 per violation per day.
Right of publicity violations create civil liability for actual damages plus profits from unauthorized use, with potential punitive damages in cases of willful violation.
The legal rules around AI generated images are real, enforceable, and actively being developed by courts, legislatures, and regulators in 2026. Most uses are legal. The exceptions are specific, serious, and increasingly well-covered by state law.
Bookmark this page as a practical legal reference before creating, selling, or publishing AI-generated images. The rules will continue to evolve, and knowing where the law stands right now is the best protection available.
