Instagram's automated "AI Content" labelling system is misfiring, tagging ordinary photos as AI-generated while genuinely AI-made images sometimes go unflagged. Users have reported the label appearing on pictures edited only with minor tools such as blemish removal or background-removal features, undermining the label's purpose of helping people quickly identify synthetic content and creating confusion about what can be trusted on the platform.

The issue echoes a similar mislabelling problem Instagram faced in 2024, when its "Made by AI" tag wrongly caught images with only minor AI-assisted edits. Meta scans for IPTC and C2PA metadata to detect AI use but has never fully explained its methods, and it did not respond to requests for clarification this time. Design platform Canva said some of its assistive AI tools, including Background Remover, were mistakenly tagged as generative and claimed to have fixed the issue, though several users say the mislabelling persists; testing by The Verge found the label only reliably appeared on content created via the Meta AI app.