Meta swaps out 'Made with AI' label after photographer backlash

Imagine digging up a 40-year-old film photo you took of a basketball game, uploading it to Instagram, and having Meta immediately slap a big badge on it that says: "Made with AI."

That exact scenario happened to former White House photographer Pete Souza. All he did was use Adobe tools to crop and flatten the image. But after weeks of furious backlash from photographers seeing their real-world work mislabeled, Meta is finally backing down.

On Monday, Meta announced it is changing the "Made with AI" tag to "AI info" across Instagram, Facebook, and Threads. Users can click the new badge for additional context on what was actually done to the image.

The label that broke photographers' trust

Meta recently broadened its detection systems to flag AI content, leaning on metadata embedded by creative software like Adobe Photoshop.

Under initiatives like C2PA, metadata acts like a digital nutrition label. It logs the creator, capture date, software used, and whether generative tools like Generative Fill were applied during the edit.

Meta's mistake was treating that data with an all-or-nothing approach. If you generated a photo entirely out of thin air, it got tagged "Made with AI." But if you took a real photo on film and used an AI-assisted retouching brush to clean a speck of dust, it also got tagged "Made with AI."

Adobe wasted no time calling out Meta's blunt implementation. Andy Parsons, senior director of Adobe's Content Authenticity Initiative, publicly pushed back, saying platforms should reserve labels like "made with AI" strictly for content that is wholly AI-generated.

Parsons noted that millions of creators use AI tools for simple aesthetic touch-ups just as they used traditional sliders for years. Slapping a fake-looking warning on legitimate work destroys creator credibility.

A cosmetic rename for now

Meta spokesperson Kate McLaughlin confirmed the new "AI info" badge is currently rolling out on mobile apps, with the web view following shortly.

Meta claims the new wording better captures that an image might simply be modified rather than generated from scratch.

Here is the catch: once you actually tap on "AI info," the pop-up modal shows the exact same copy as before. It still broadly lumps fully synthetic images together with basic photo edits.

In other words, Meta tweaked the optics to stop the outrage, but hasn't actually solved the underlying nuance problem.

Why it matters

For photographers and digital artists, this change cools the temperature, but the risk of confusion remains. Until platforms display a clear breakdown of edits, your manual work is still at the mercy of opaque metadata flags.

We think this mess highlights the growing pains of digital provenance. Standards like C2PA are designed to provide granular context, but platforms like Meta flattened that rich data into a binary scarlet letter.

Meta took the lazy route to appease regulators and curb synthetic media, and real creators became collateral damage. If the industry wants consumers to trust digital media, platforms have to learn the difference between using an AI tool to touch up reality and generating a synthetic hallucination from scratch.