Microsoft Sets Rules for AI-Generated Advertising

Microsoft is strengthening its requirements for transparency around the use of artificial intelligence in advertising. The company requires advertisers to disclose information about synthetic content and preserve data about its provenance.

Using artificial intelligence to create advertising materials is becoming common practice. At the same time, Microsoft Advertising emphasizes that the use of AI does not exempt advertisers from requirements related to transparency, obtaining the necessary permissions, and preventing users from being misled.

Microsoft Publishes Rules for Using AI in Advertising

Microsoft Advertising has published dedicated guidance on the use of content created with artificial intelligence, AI-manipulated content, and other types of synthetic content in advertisements.

Advertisers are responsible for complying with applicable laws in the markets where their ads are displayed. This includes requirements related to AI disclosures, obtaining necessary permissions, and using a person’s image, voice, or other identifying characteristics.

Therefore, using AI to create an advertising creative is not itself a violation of the rules. However, advertisers must comply with transparency and legal requirements applicable to such content.

What Information About AI Content Should Be Preserved

Microsoft recommends that advertisers preserve watermarks, metadata, and other provenance information that can help identify how content was created using artificial intelligence.

Such information can be used to verify the origin of images, videos, audio, and other advertising materials.

When users must be informed that AI was used to create specific content, advertisers are also responsible for providing that disclosure appropriately. Microsoft recommends integrating AI disclosures directly into images or videos.

For supported formats, advertisers can also use Microsoft’s existing ad disclaimer feature.

An “AI-Generated” Label Does Not Make Deceptive Advertising Acceptable

Adding a label indicating that AI was used does not automatically make an advertising creative compliant with Microsoft’s requirements.

The company states that ads may be rejected, restricted, or removed if they contain prohibited deepfakes, impersonate people or organizations, use a person’s image or voice without the required authorization, fail to include mandatory disclosures, or interfere with machine-readable provenance information.

This means that AI disclosure is only one part of the requirements for advertising materials. It does not make content acceptable if it violates other platform rules or applicable legal requirements.

How Microsoft Handles Its Own AI-Generated Content

Microsoft also explains how content created using its own artificial intelligence tools is handled.

Images, audio, and video generated with Microsoft’s AI tools may contain machine-readable provenance information, metadata, and imperceptible watermarks indicating that artificial intelligence was used.

However, these technical signals are not necessarily visible to consumers. As a result, even when AI-generated content already contains relevant metadata or a watermark, advertisers may still need to provide an additional visible or audible disclosure about the use of AI.

What Advertisers Should Check

Microsoft recommends reviewing advertising materials created or modified using AI before they are submitted for publication.

In particular, advertisers should make sure that the people, products, places, claims, and events depicted in the creative are accurate.

Special attention should be paid to content that could create a false impression about real products, services, people, or events. Using generative AI does not remove an advertiser’s responsibility for the content of an ad.

Microsoft also recommends avoiding the removal of metadata and watermarks from AI-generated assets when they contain information about the content’s provenance.

Advertisers should also check disclosure requirements in every country where a campaign will run. Rules concerning the use of synthetic content, people’s images, and voices may vary between markets.

What This Means for Advertisers

Microsoft is not banning the use of artificial intelligence to create advertising. Instead, the company is establishing clearer requirements for how advertisers should work with AI-generated content.

Using AI to create or edit advertising materials should involve appropriate controls. Advertisers need to verify the accuracy of the content, preserve provenance information, obtain the necessary permissions, and comply with disclosure requirements.

For advertisers, this means that the process of preparing AI-generated creatives should involve more than simply generating the materials. The content should also be reviewed before a campaign is launched.

Microsoft is effectively drawing a line between using artificial intelligence as a tool for creating advertising and using AI to mislead consumers. AI-generated creatives can be used in advertising, but they must comply with requirements for transparency, accuracy, and permitted advertising content.

Read this article in Ukrainian.