After saying it was evaluating adjusting its policy last week, Meta today announced that it is making changes to how it handles what it calls “manipulated media” going forward, adjusting the “Made with AI” label that appears on Instagram to “AI Info” instead.
Meta made these changes earlier this year in response to recommendations from its Oversight Board, which said that providing “transparency and additional context is now the better way to address manipulated media and avoid the risk of unnecessarily restricting freedom of speech.”
However, Meta’s first approach — the “Made with AI” tag — was broad, heavy-handed, and often inaccurate PetaPixel reported. There are a host of tools inside of Adobe Photoshop that will generate the “Made with AI” tag even though the final image may be 99% or more untouched by AI.
While Meta is changing the wording, it is not deviating entirely from its efforts to label what it deems is AI-generated content.
“We want people to know when they see posts that have been made with AI. Earlier this year, we announced a new approach for labeling AI-generated content. An important part of this approach relies on industry-standard indicators that other companies include in content created using their tools, which help us assess whether something is created using AI,” Meta writes in an update to its blog post.
“Like others across the industry, we’ve found that our labels based on these indicators weren’t always aligned with people’s expectations and didn’t always provide enough context. For example, some content that included minor modifications using AI, such as retouching tools, included industry standard indicators that were then labeled ‘Made with AI.’ While we work with companies across the industry to improve the process so our labeling approach better matches our intent, we’re updating the ‘Made with AI’ label to ‘AI info’ across our apps, which people can click for more information.”
Meta’s examples of the new tag — above — appear next to what appear to be fully AI-generated images, which has never been the source of the backlash.
The issue with Meta’s approach is that it will still seem to “punish” photographers who use Adobe Photoshop more than those who use full generative AI programs, like Stable Diffusion, since the “industry standard indicators” it cites are not universally applied. These indicators, while not specifically defined by Meta, are likely the ones that Content Authenticity Initiative (CAI) partners are using. Meta is, notably, not a member of the CAI.
Image credits: Meta