After Meta introduced the “Made with AI” label for photos in May, photographers voiced concerns that the label was incorrectly applied to photos where basic editing tools were used. Responding to user feedback and to clarify the extent of AI involvement in photo creation, Meta is now transitioning the label to “AI Info” across all its apps.
Meta acknowledged that the previous label didn’t sufficiently clarify that photos labeled “Made with AI” could include minor edits using AI-powered tools like retouching. In an updated blog post, Meta explained, “Like others in the industry, we found that our labels based on these indicators sometimes didn’t align with people’s expectations and lacked context.”
The underlying technology for detecting AI use in photos remains unchanged, leveraging standards such as C2PA and IPTC to identify AI tools utilized in editing processes. For example, photos modified with tools like Adobe’s Generative AI Fill may still receive the new “AI Info” label. Meta aims for the updated label to better communicate that the presence of AI tools does not necessarily mean the entire photo was AI-generated.
Kate McLaughlin, a spokesperson for Meta, emphasized that “‘AI Info’ can encompass content that was made and/or modified with AI, aligning more closely with user expectations.” Meta plans to collaborate with industry peers to refine these labeling practices further.
While the new tag clarifies AI involvement in photo editing, it does not address the detection of entirely AI-generated photos or specify the extent of AI-powered edits in labeled images. Moving forward, Meta and other social networks will need to establish clear guidelines that balance transparency with fairness to photographers using AI-powered tools for editing.
Source: Meta