Users reported that Meta automatically added an “AI content” tag to images that were not created or edited using generative AI tools.
A report by “The Verge” website stated that this mechanism has raised increasing criticism over the past weeks, after users noticed that the platform automatically places the tag on images that were not created or modified with generative artificial intelligence tools.
According to the site, there are multiple reasons that may lead to the appearance of this mark, as users reported that it appeared on images that were modified using simple tools, such as the background removal tool in “Canva,” or tools to slightly correct defects.
One user of the “Threads” platform said that “Instagram” places the mark “every time the background removal tool is used,” while another user said that it appeared after he used “Canva” to remove a “small spot” from his photo, which led to the entire image being classified on “Instagram” as modified by artificial intelligence.
Users’ doubts increased about Instagram’s ability to distinguish between original, modified, and artificially intelligent content, especially after images generated using artificial intelligence appeared without the tag.
Meta relies on metadata to monitor content modified with artificial intelligence, amid questions about the accuracy of this mechanism, especially after cases showed that using auxiliary tools, such as removing the background, may lead to images being labeled as generated by artificial intelligence. (Erm News)
