Google Messages may soon flag AI-altered photos in your chats 🔍

Cliquez ici pour lire en français

AI has already reshaped how we write, search, and share. Now it’s blurring a line that cuts even deeper: the one between a real photo and a fabricated one. Google Messages, it turns out, may be building a tool to help you tell the difference right inside your conversations.

A feature built to settle the doubt 🔍

Code spotted in a recent beta build suggests Google Messages could soon tell you whether an image was generated, edited, or combined using AI tools. The idea isn’t just to slap an « AI » sticker on a photo — it’s reportedly meant to surface a more granular read on where that image actually came from.

It sounds like a minor, almost technical tweak. But it touches something very real: how much we trust what lands in our inbox every day. In a private chat, a fake photo might just get a laugh. In a wider context, the same image can mislead, manipulate, or help a false story spread.

C2PA: the standard Google is reportedly betting on 🧩

The feature reportedly leans on C2PA — the Coalition for Content Provenance and Authenticity — an industry standard built to read metadata and trace a piece of content back to its origin. Based on the strings found so far, Google Messages could surface labels along the lines of « edited with multiple AI tools, » « unedited capture, » or « parts of this media were made with AI. »

What makes this interesting is that it isn’t just trying to answer a yes-or-no question. It’s trying to tell the image’s story. And in a world where a convincing fake can be generated in seconds, that paper trail is starting to matter almost as much as the image itself.

Trust becomes an interface ✨

If this feature actually ships, Google Messages would be nudging users toward a small but important habit: check where an image came from before you believe it. In a feed saturated with generated content, that one prompt won’t fix everything — but it gives our conversations back a bit of clarity.

The real question here probably isn’t whether AI is everywhere — at this point, it is. It’s when our everyday apps will finally start helping us tell the authentic from the fabricated.

Does an AI label in your messaging app actually rebuild trust — or are we going to need a lot more than that?