Making an ad used to mean a photo shoot, a video crew, or at least a free afternoon. Now you can type a sentence and get a polished product photo or a short video clip in minutes.
That is a real advantage, especially for small teams. But AI visuals can also backfire. A customer who spots a fake photo of “your team” or a too-perfect “finished project” may stop trusting everything else you say. Here is where AI images and AI video ads help, where they hurt, and what the major platforms now ask you to disclose.
Where AI visuals work well
AI is great for the frame around your message. Good uses include:
- Product photos on new backgrounds, like a seasonal or lifestyle scene
- Several versions of one ad, so you can test which image or hook works best
- Simple illustrations and icons for blog posts and social graphics
- Short animated explainers that show how a service works, step by step
- Storyboards and rough drafts before a real shoot
- Background footage that sets a mood without making a claim
In each case, the AI image does not pretend to be proof of anything. It supports the message instead of standing in for reality.
Where AI visuals hurt trust
Problems start when an AI image is used as evidence. Avoid AI for:
- Before-and-after photos. If a contractor, dentist, or cleaning company shows AI-made results, it is misleading.
- “Our team” or “our office” photos. Customers expect to meet the people they see.
- Testimonials from people who do not exist. The FTC’s rule on fake reviews, in effect since October 21, 2024, bans fake testimonials, including ones made with AI.
- Food, rooms, or products that look better than what the customer will actually get.
- Anyone’s face or voice without permission, including celebrities.
A simple test: if a customer would feel tricked after learning the image was AI-made, use a real photo.
What the platforms ask you to disclose
The rules are moving fast, but here is where the major ad platforms stand as of October 2026.
Google Ads
In July 2026, Google added a “How this ad was made” section to the My Ad Center panel for ads on Search, YouTube, and Discover. Ads built with Google’s own AI tools get the disclosure automatically. If you make assets with outside AI tools, Google gives you a control to label them yourself. In some places, including the EU, India, and New York, a label can also appear on the ad itself.
Meta (Facebook and Instagram)
Meta adds an “AI info” label to ads that were created or significantly edited with its own generative AI features, such as background generation or image generation. It can also label images made with outside AI tools when it detects them. Ads that show photorealistic AI-generated people may get a more visible label.
YouTube
YouTube requires creators to disclose realistic altered or synthetic content, such as making a real person appear to say something they did not, or showing a realistic scene that never happened. You do not need to disclose clearly unrealistic content, animation, or AI used for things like scripts and captions.
An AI label is not a penalty. The platforms have not published data showing that labeled ads perform worse, and being upfront tends to protect trust. Test it with your own audience.
A simple policy for your business
Write down a few rules your team and any vendors can follow:
- Real work, real people, real results: always use real photos and video.
- AI is fine for backgrounds, concepts, graphics, and drafts.
- Label realistic AI images and video when a platform asks, and whenever a customer could be misled.
- Keep the original prompts and files so you can show how an asset was made.
- Check every AI asset for odd hands, warped text, wrong logos, and details that do not match your real business.
This keeps AI doing what it does well, which is saving time, while your real work does what it does best, which is earning trust. It also fits how strong social media marketing and Google Ads campaigns work today: fast testing on top of honest proof.
Your website deserves the same care. Real project photos on your service pages give visitors confidence and help your website turn visits into calls.
Frequently asked questions
Are AI-generated ads allowed on Google and Facebook?
Yes. Both allow AI-made ad creative as long as it follows their normal ad policies and disclosure rules. Misleading ads are not allowed, whether AI made them or not.
Do I have to label AI images on my own website?
There is no single US rule that covers every image on every website. But deceptive marketing is illegal no matter how it was made, and some places have their own AI rules. When in doubt, be open about it.
What are the best AI video tools for ads?
Tools change quickly. Pick one that gives you commercial usage rights, lets you edit the results, and adds content credentials or watermarks that show how the media was made.
Sources
- FTC: Final rule banning fake reviews and testimonials (August 2024)
- Search Engine Roundtable: Google Ads now show AI creation labels (July 2026)
- NBC News: YouTube will require creators to label realistic AI content
About the author
Vishal Chandra, Head of Operations at Brandmasterz. Vishal has 7 years of experience in marketing and oversees how client work gets planned, delivered, and reported at Brandmasterz. Outside work, he plays guitar and performs with his band.