What is the "How this ad was made" disclosure?
"How this ad was made" is a section in Google's My Ad Center that can provide users with information about whether assets were created or edited using generative AI. Google says users can access the information through the three-dot menu on ads across Search, YouTube, and Discover.
When should advertisers label AI-generated or AI-edited assets?
Label assets when they were created or materially edited with generative AI and the platform workflow requires or supports disclosure. Keep a production record so the media team knows which assets need designation.
Examples likely to require review:
- –An image generated entirely from a prompt
- –A product scene generated around a real product
- –A background replaced with generative AI
- –People or objects added or removed
- –Material visual elements reconstructed
- –Video generated or materially altered with AI
What can Google label automatically?
Google says it may label assets in some cases, including where it is legally required or where signals from other platforms indicate AI use. When advertisers use fully automated Google features to create and serve creative, Google may apply labels on their behalf. Google also says labels applied in these cases may not be overridden.
Where can visible AI labels appear?
Google's current documentation says the My Ad Center disclosure is globally accessible. It also identifies some geographies where designated AI-created or edited assets can receive visible overlays on the ad itself, including campaigns targeting the European Union, India, and New York. Because requirements can change, verify the current geography list before launch.
What should an internal AI creative register include?
| Field | Example |
|---|---|
| Asset ID | img-pmax-summer-014 |
| Campaign | PMax Summer 2026 |
| Source | Brand photo + AI background edit |
| Tool | Google Asset Studio |
| AI created or edited | Edited |
| Human reviewer | Name/role |
| Product accuracy checked | Yes |
| Legal/policy checked | Yes |
| Label setting | AI edited |
| Approval date | 2026-07-12 |
What should be reviewed before an AI ad asset is published?
| Risk | Review question |
|---|---|
| Product truth | Does the asset show the actual product, size, color, packaging, and functionality? |
| People | Does it create a misleading person, testimonial, endorsement, or demographic implication? |
| Location | Is the environment real, illustrative, or potentially misleading? |
| Claims | Does copy imply unsupported performance, safety, savings, or results? |
| Brand | Does it match brand identity and quality? |
| Rights | Does the advertiser have rights to source assets and outputs? |
| Disclosure | Is the correct AI label setting applied? |
| Policy | Does the asset comply with Google Ads and industry rules? |
AI creative governance workflow
Common mistakes
- –Not asking whether client assets used AI
- –Publishing generated products that do not match reality
- –Using synthetic testimonials or fake endorsements
- –Ignoring geographic disclosure requirements
- –Assuming Google will label everything automatically
- –Assuming an advertiser-applied label can override a Google-applied label
- –Using AI to create volume without quality control
Related guides
Source notes
- Google Ads Help, "Use AI content label settings and disclosures," updated July 2026.
- PPC News Feed reporting on expanded Google Ads AI transparency features.
- Google Ads policies and creative asset documentation.