Can Your Business Use AI-Generated People In Advertising?

So, it’s time to get your business’s marketing sorted. Instead of organising a photoshoot, hiring a photographer, getting release forms signed and dealing with everything else that comes with creating new advertising content, you’ve decided to use AI.

You’ve probably seen other businesses doing the same. AI-generated people can look surprisingly realistic, and creating an image can be much quicker and cheaper than organising a traditional photoshoot. So, why not?

Businesses can generally use AI-generated people in advertising, but AI-generated does not mean risk-free or rights-free.

Before publishing the image, you still need to consider where it came from, whether you have permission to use any material or real people involved, what the AI provider allows and what the finished advertisement leads customers to believe.

Disclosure can also be relevant, but it is only one part of the legal picture.

What Counts As An AI-Generated Person?

Not every AI-generated person is created in the same way.

At one end, you might type a prompt into an AI tool and generate a completely fictional person who was not intentionally based on anyone real.

However, you might also upload an existing photograph and ask an AI tool to create something new from it. This could involve changing the person’s clothing, placing them in a new setting or generating a series of advertising images from a single photograph.

AI can also be used to recreate or closely imitate the appearance of an identifiable real person.

The important distinction is whether you are working with a fictional person or using the image, identity or likeness of someone who already exists.

If the person is completely fictional and the image does not reproduce someone else’s protected work, some of the consent and privacy issues discussed below may be less relevant.

Once an existing photograph or identifiable real person is involved, you need to look more closely at what rights and permissions your business actually has.

Does AI-Generated Mean Copyright-Free?

No. The fact that an image was created using AI does not automatically remove copyright issues.

One of the first things to check is what your business gave the AI tool.

For example, you might upload an existing photograph and ask the tool to place the person in a different setting, change their clothing or create several new versions of them for a campaign.

Before doing that, consider who actually owns the photograph.

The person appearing in a photograph does not necessarily own copyright in it. Copyright may instead belong to the photographer, agency or another party. If your business licensed the image, the licence may also limit how the photograph can be altered, reproduced or used commercially.

Permission to use an image on your website, for example, does not necessarily mean you have permission to upload it to an AI system and use it to create new advertising.

The practical solution is to check the licence or agreement covering the original material before uploading it.

If your business regularly commissions photography or other creative work, your contracts can also deal clearly with ownership, editing rights, commercial use and, where appropriate, AI-assisted modification.

If you are unsure whether your existing rights cover a particular use, a Copyright Consult can help clarify what your business can legally do with the material.

There is also the finished AI output to consider.

An image does not become immune from copyright claims simply because an AI system generated it. If the output reproduces protected elements of an existing photograph, artwork or other work closely enough, copyright issues may still arise.

So, if an AI-generated image looks unusually similar to an existing photograph, artwork or advertising campaign, it is worth investigating that similarity before publishing it rather than assuming the AI provider has cleared the rights for you.

What About Copyrighted Material Used To Train AI?

There is also a broader legal debate around the copyrighted material used to train generative AI systems.

AI models can be trained on very large collections of images and other material, and governments, courts, copyright owners and AI companies around the world are continuing to work through when and how copyright material can lawfully be used for AI training.

Australia has not settled every part of this issue. The Attorney-General’s Department is currently considering issues including licensing copyright material for AI, greater certainty around material generated using AI and enforcement relating to AI outputs. Australia is also not currently proposing a broad text and data mining copyright exception. See the Copyright and Artificial Intelligence Reference Group for more information.

This does not mean every image produced by an AI system infringes copyright. A generated face is not automatically a copy of one particular photograph simply because existing images may have formed part of a model’s training data.

For a business using an existing AI tool, the more immediate checks are usually simpler: know what you uploaded, check the rights you have over it, understand the provider’s terms and review the finished image before publishing it.

Can Your Business Own The AI-Generated Image?

There is another side to the copyright question: if your business creates an AI-generated advertising image, can you stop someone else from copying it?

The answer may depend on how much human creative input went into the finished work.

IP Australia notes that copyright protection for many types of works requires a human author, and the extent to which AI has been used can create difficult questions about whether copyright protection applies.

Its guidance recommends keeping records of what your team planned, the prompts used and what was selected or edited by a person where IP protection is important. You can read more in IP Australia’s guidance on using AI to create IP.

This can matter if an AI-generated character or campaign is going to become an important business asset.

If you are investing heavily in a campaign and exclusivity matters, it can be worth thinking about ownership before building the entire campaign around an image that may be difficult to protect later.

What Do The AI Tool’s Terms Let You Do?

Before using an AI-generated image commercially, check that the tool actually permits the way you intend to use it.

Different AI providers have different terms around uploaded material, generated outputs, commercial use and responsibility if someone else claims their rights have been infringed.

In particular, check:

• whether commercial use is permitted
• what rights you receive over the output
• whether there are restrictions on recreating or generating real people
• what happens to photographs and other material you upload
• whether your inputs can be stored or used to train or improve the service
• who carries the risk if a third party claims the content infringes their rights.

There is an important distinction here.

An AI provider giving you permission to use its output commercially does not necessarily mean the provider is guaranteeing that nobody else has rights in the image.

If your business will be relying heavily on a particular AI platform and the terms are unclear, having the agreement reviewed before relying on it commercially can help you understand what you are actually agreeing to.

Businesses should also think about who within the team is allowed to use AI tools.

If employees are generating social media posts, advertisements or other customer-facing material, an Employee AI Use Policy can set rules around approved tools, what staff can upload, IP and privacy, and when AI-generated content needs human approval before publication.

What If The AI Person Is Based On Someone Real?

If a real person is identifiable, do not assume an old photo permission automatically covers creating new AI versions of them.

Say an employee agrees to appear in a photograph for your website.

A few months later, you upload that photograph to an AI tool and create dozens of new versions of them for social media advertisements, videos and product campaigns.

That is a broader use than simply publishing the original photograph, and the permission you originally received may not cover it.

There can also be separate rights involved.

Copyright in the original photograph is one question. Whether you can use the individual’s identity, likeness or personal information in this new way is another.

For businesses covered by the Privacy Act 1988, the OAIC says privacy obligations can apply to personal information both input into an AI system and generated through it. Its guidance also warns businesses to consider whether putting existing personal information into an AI product amounts to a new use or disclosure. You can read more in the OAIC’s guidance on commercially available AI products.

This does not mean every AI image involving a person automatically breaches privacy law. However, if your business is taking an identifiable person’s information and using it in a new AI workflow, privacy should be part of the check.

There may also be an advertising issue if using somebody’s likeness gives customers the impression that the person approves of, endorses or is affiliated with your business when they are not. The Australian Consumer Law prohibits certain false or misleading representations about sponsorship, approval and affiliation.

The safer approach is to get clear permission for what you actually intend to do.

If an employee, customer, model or creator will appear in AI-generated marketing, a suitable Talent Release Form or other written agreement can set out how their image, voice or performance can be used and whether editing or AI-generated versions are permitted.

If your business is also handling personal information through AI systems, make sure your Privacy Policy and relevant collection notices accurately reflect what the business is actually doing. A Privacy Policy is not a substitute for consent where consent is required, but it is an important part of being transparent about your data practices.

Can An AI-Generated Person Give A Testimonial Or Endorse A Product?

An AI-generated person should not be presented as having a genuine customer experience that never happened.

Using an AI-generated person to hold your skincare product is one thing.

Having that person say:

“I’ve used this for six weeks and my skin has completely cleared up.”

is another.

The second advertisement suggests that someone has actually used the product and experienced those results.

Under the Australian Consumer Law, businesses must not make false or misleading representations that purport to be testimonials, or false or misleading representations concerning testimonials. It also deals with false representations about sponsorship, approval, performance, benefits and affiliation.

The ACCC also makes it clear that reviews should reflect genuine experiences and that businesses must not create fake or misleading reviews. You can read more in the ACCC’s guidance on online reviews.

This does not necessarily mean an AI avatar can never talk about your product.

The safer distinction is between using AI as a presenter and inventing an experience, qualification or endorsement that does not exist.

If an AI character is simply presenting factual information about your product, make sure those claims are accurate and can be supported.

If you want to use a testimonial, customer story or endorsement, use a genuine one and make it clear whose experience it actually represents.

The same care should be taken with AI-generated doctors, lawyers, tradespeople or other apparent professionals. If an advertisement creates the impression that a qualified person has approved, recommended or personally used a product when nobody has, that can create a misleading impression.

Can The AI Image Misrepresent Your Product?

Yes. The image itself can make a claim, even if the advertisement says very little.

For example, AI could be used to create:

• a before-and-after image for a skincare product
• a model wearing clothing that has been digitally altered to fit differently
• someone demonstrating a product achieving a result it cannot actually achieve
• an AI-generated customer showing a particular health, fitness or beauty result.

Even without a written claim, those images can create an impression about what your product looks like, what it does or what results customers can expect.

The ACCC says claims about products must be accurate and truthful, and specifically includes images and descriptions as information that can create a false impression. See the ACCC’s guidance on false or misleading claims.

The practical check here is simple: compare the AI-generated advertisement with the real product before you publish it.

If AI has changed how the product looks, works or performs in a way that could give customers the wrong impression, change the advertisement rather than relying on a small disclaimer to correct it.

Do You Have To Disclose That The Person Is AI-Generated?

Australia does not currently have a blanket law requiring every AI-generated advertising image to be labelled as AI-generated.

However, that does not mean disclosure is never appropriate.

The National AI Centre’s guidance on being clear about AI-generated content encourages businesses to take a proportionate approach to transparency. It says AI use may not need to be disclosed every time, but businesses should consider factors such as how much AI contributed to the content and whether it could affect people’s decisions, rights, safety or trust.

Suggested transparency methods include labels, watermarks and metadata.

For an advertiser, a useful question is:

Would a customer understand this advertisement differently if they knew how the person or image was created?

If an AI-generated person could reasonably be mistaken for a genuine customer, employee or expert, disclosure may help avoid the wrong impression.

However, transparency should not be treated as a cure-all.

Writing “AI-generated image” underneath an advertisement does not automatically:

• give you copyright permission to use an existing photograph
• give you permission to create an AI version of a real person
• turn an invented testimonial into a genuine one
• make an inaccurate product claim acceptable
• override the AI provider’s terms.

Disclosure deals with transparency. It does not create rights that your business did not already have.

What If Your Advertising Is Seen Overseas?

If your business advertises internationally, complying with Australian requirements may not be enough.

AI regulation is developing differently around the world, and the laws of the markets you advertise into may also apply depending on the circumstances.

For example, transparency requirements under Article 50 of the European Union’s AI Act began applying on 2 August 2026. These include obligations for providers of certain AI systems to make AI-generated or manipulated content detectable using machine-readable marking, as well as disclosure obligations where AI is used to create or manipulate content that constitutes a deepfake.

This does not mean that every AI-generated advertising image used in Europe simply needs the same visible “AI-generated” label. The requirements depend on the type of content and the role of the business or AI provider.

The European Commission’s Article 50 guidance explains these requirements in more detail.

For Australian businesses, the practical point is simple: if you run the same campaign across several countries, do not assume the rules that apply in Australia will be enough everywhere else.

This is particularly important where a campaign uses realistic synthetic people, recreates an identifiable person or contains content that customers could reasonably believe is authentic.

What Should You Check Before Using An AI Person In An Ad?

AI can make creating advertising content faster and easier, but it should not mean skipping the checks you would normally make before a campaign goes live.

Before publishing an AI-generated person, ask:

• Is this person completely fictional, or could they be identified as someone real?
• Did we upload an existing photograph, artwork or other third-party material?
• Do we have the copyright or licence needed to use and modify it?
• If a real person is involved, does their permission actually cover AI modification and advertising?
• Do the AI tool’s terms allow this type of commercial use?
• Does the finished image look unusually similar to an existing creative work?
• If this campaign becomes valuable, do we understand what rights we have in the AI-generated output?
• Could the person appear to be a genuine customer, employee, expert or endorser?
• Are we presenting any experience or testimonial as genuine when it is not?
• Does the image accurately represent the product and the results it can achieve?
• Would disclosing the use of AI make the advertisement clearer?
• Where will the advertisement appear, and do different rules apply in those markets?

You do not necessarily need legal advice every time your marketing team generates an AI image.

However, if you are using a real person’s likeness, uploading third-party creative material, relying on unclear AI-provider terms or preparing a significant campaign for several countries, speaking to a legal expert before publishing can help you work out what permissions, contracts or changes you actually need.

AI might make creating the advertisement easier. The important part is making sure it does not also create a copyright, privacy or advertising problem that could have been avoided with a few checks upfront. 

If you would like a consultation on using AI generating people in your business’s advertising, you can reach us at 1800 730 617 or team@sprintlaw.com.au for a free, no-obligations chat.