Meta is enhancing transparency in advertising by labeling ads created or edited with generative AI tools. This initiative, developed in collaboration with experts, aims to clarify the influence of these tools on ads, thereby improving trust. Labels will indicate significant edits or the inclusion of AI-generated photorealistic humans, with plans to expand this labeling to non-Meta AI tools as expectations evolve.
Meta prioritizes transparency and responsibility in its advertising practices. Since last year, they have been labeling ads created or significantly edited using their generative AI features available in advertiser marketing tools. This approach is informed by ongoing collaboration with experts and stakeholders, recognizing the rapidly changing transparency expectations surrounding generative AI.
AI Labeling System
The AI labeling system is designed to inform users when images and videos have been created or significantly edited with Meta's generative AI tools. Labels will appear in the three-dot menu or next to the “Sponsored” label based on the extent of AI involvement:
- No Label: If the generative AI tools do not result in significant edits or include photorealistic humans.
- AI Label: If significant edits are made or if an AI-generated photorealistic human is included, the label will appear next to the Sponsored label.
Meta is committed to refining its ads transparency tools and will share more information on labeling for non-Meta generative AI tools. This effort is part of a broader goal to build trust and accountability in advertising practices. Users can still access information about why they see specific ads through the “Why am I seeing this ad?” tool, hide ads from specific advertisers, and manage their ad preferences. Meta emphasizes responsible use of machine learning models and generative AI in delivering ads.
