Meta Revises AI Manipulated Content Labels To 'AI Info' On Social Platforms
Meta has revised its method of labeling AI-manipulated content on platforms like Instagram and Facebook. The company initially posted a blog in April outlining its plans to label AI-generated content, and has now updated the "Made with AI" label to "AI info" across its apps. Unlike the previous static label, the new AI info badge will be clickable, providing more information about the content.
There have been reports of Meta mistakenly labeling real photographs taken by photographers as "Made by AI." For instance, a basketball game photo by Pete Souza, a former White House photographer, was marked as AI-generated. Meta explained that it uses "industry standard indicators" provided by AI companies to identify content created using their tools. However, even minor edits to images are being detected by this mechanism, resulting in incorrect labeling.
Meta acknowledged that its previous labels based on these indicators did not always align with people's expectations and lacked sufficient context. The company stated that even minor AI modifications, such as retouching tools, were labeled as "Made with AI" due to the presence of industry standard indicators.
To address this issue, Meta is collaborating with industry partners to enhance its labeling approach. The updated "AI info" label is a result of these improvements. Meta plans to apply the new label to a broader range of video, audio, and image content when it detects industry standard AI image indicators or when users disclose that they are uploading AI-generated material.
Sign in with Google