Calabi Labs · Guide · 2026-06-13
Yes, Meta is expanding its AI-image labeling efforts on Facebook and Instagram ahead of the November 2024 election. The company announced in February 2024 that it would significantly broaden detection of AI-generated content across its platforms, with labels rolling out starting in May 2024.
Meta, which owns Facebook and Instagram, is expanding its ability to identify and label AI-generated images posted on its platforms. The company says it can now detect many AI-made images using industry-standard indicators—metadata signals embedded in AI-generated content from tools like Adobe Firefly, Midjourney, and other popular image generators.
These "Made with AI" labels are being applied to:
The labels appear on content posted to Facebook, Instagram, and Threads, alerting users when content has been created with artificial intelligence rather than captured by a camera or created by hand.
The timing is deliberate. With major elections scheduled globally—including the U.S. presidential election in November 2024—Meta is under pressure to combat AI-generated misinformation and deepfakes that could spread on its platforms. The company has faced criticism for inadequate moderation of manipulated media in past elections.
Meta's expanded labeling effort is part of a broader strategy to increase transparency around AI-generated content, giving users more context to evaluate what they see in their feeds.
Meta's detection system looks for industry-standard metadata markers that AI image generators increasingly embed in their output. When these markers are present, Meta can automatically apply the "Made with AI" label. The company has also been working to improve detection of AI-generated content that lacks such markers, using behavioral signals and other techniques.
If you browse Facebook or Instagram and see a "Made with AI" label on an image, it means Meta's systems detected AI-generation markers in that content. This doesn't necessarily mean the content is misleading—it's simply transparency about how the image was created. However, the labels are intended to help users identify potentially manipulated content, especially around election-related topics.
Try Calabi free at calabilabs.com — 10 cleans, no card.