Meta replaces 'Made with AI' label on Instagram following complaints
What's the story
Meta, the parent company of Instagram, has decided to replace its 'Made with AI' label with 'AI info' across its applications.
This decision was prompted by numerous complaints from photographers, including former White House photographer Pete Souza, who reported their images being inaccurately tagged as 'Made with AI.'
The issue stemmed from how platforms interpreted metadata tools like Adobe Photoshop applied to images.
Problem
Mislabeling issue traced to Adobe Photoshop metadata
The mislabeling problem arose when Meta expanded its AI content labeling policies, causing real-life images posted on Facebook, Instagram, and Threads to be tagged as 'Made with AI.'
Souza discovered this issue when a photo he took 40 years ago at a basketball game was incorrectly labeled.
He speculated that the use of Adobe's cropping tool and image flattening might have triggered the incorrect tag.
Statement
Meta's response to labeling concerns
Meta spokesperson Kate McLaughlin stated, "As we've said from the beginning, we're consistently improving our AI products, and we are working closely with our industry partners on our approach to AI labeling."
The new 'AI info' label is intended to more accurately represent content that may have been modified using AI, rather than suggesting it was entirely generated by artificial intelligence.
Cause
'Made with AI' label triggers on minor modifications
Meta introduced the 'Made with AI' labels earlier this year in response to criticism about its "manipulated media" policy.
However, photographers started noticing that Facebook and Instagram were applying the badge to images that hadn't been created with AI.
Tests conducted by PetaPixel revealed that even minor edits done via Adobe's generative fill tool in Photoshop, could trigger the 'Made with AI' label.