Meta refines AI labelling in ads amid growing concerns over misinformation

Meta is enhancing its artificial intelligence (AI) transparency efforts, refining disclosure labels for ads created using its in-house generative AI tools. The company initially introduced these labels last year and continues to improve them to ensure clarity for users.

According to Meta, the updated labels will appear either in the three-dot menu or next to the 'Sponsored' tag when an image or video has been 'significantly' altered using its AI tools. However, the company has not explicitly defined what constitutes a 'significant' edit.

“If an advertiser is using our in-house generative AI creative features and these tools do not result in significant edits to the image or video and do not include a photorealistic human, then we will not apply any AI labels,” Meta stated.

The current labelling system applies only to paid advertisements, leaving a gap in transparency for AI-generated content in organic posts. AI-created images are increasingly engaging Facebook users, often without clear disclosure, raising concerns over misinformation and deceptive content.

Meta acknowledges this issue and is working on a broader approach to labelling AI-generated content, including images created through third-party AI tools.

“This year, we also plan to share more information on our approach to labelling ad images made or edited with non-Meta generative AI tools,” the company noted.

While Meta pushes for more AI integration across its platforms, critics argue that the proliferation of AI-generated content could degrade user experience. Despite growing concerns over misinformation, the company continues to promote its AI chatbot and image-generation features, aiming to establish itself as a leader in AI-powered content.