US social media giant Meta Platforms Inc. is implementing a minor modification to its artificial intelligence (AI) label to resolve an issue that mistakenly identified certain images uploaded on its social media platforms as AI-generated.
The Facebook parent company is changing its “Made with AI” tag, introduced in February, to “AI info” across all its mobile apps, including its flagship platform, Instagram, and Threads.
The new label will now only indicate the presence of AI-enabled tools in creating or editing a photo, providing a brief description of how AI could have been involved in the image’s composition.
The main technology of the AI detection system will be as it is, according to the firm, suggesting the new tag could be added to photos created or modified with editing software tools such as those of Adobe Inc.
AI Risks Eliminating Divide Between Real and Not
Meta encountered significant online criticism in late June, with multiple photographers saying their photos had been inaccurately tagged as “Made with AI” when viewed on mobile.
Minor retouching on real images has been mistakenly labeled as AI-generated, such as when using Adobe’s Generative AI Fill to remove small objects from the photo.
The “AI info” change is expected to clear the confusion for users, particularly the artists and photographers whose works have been incorrectly labeled as entirely developed by AI despite them only applying slight changes to their creations.
While the issue has now been fixed, the incident signals the challenges the growing AI landscape may pose in perceiving real work from artificial and sentiment over the creative skills of a human against a machine.
There is currently major controversy around AI-produced art, which uses text prompts, as the potential of such becoming mainstream could risk eliminating the space separating the real from the unreal.
The focus now is on providing more in-depth information about AI technology’s role in visual content development and its possible impact on creative industries.