Real or AI

How to tell if a photo was made with AI

Updated 14 November 2026

हिन्दी · Español · Português

AI image tools such as ChatGPT, Gemini, Midjourney and FLUX now make photos that look like they came from a phone. The old tricks (count the fingers, look for melted text) catch fewer of them every month. This guide covers what still works, what doesn't, and how to check a photo quickly.

Check a photo now

1. Look at the details nobody checks

AI tools get the main subject right and slip on the background. Zoom in on:

The newest tools make far fewer of these mistakes. If you find one, it is a strong sign; if you find none, that proves nothing.

2. Ask where the photo came from

3. Check the photo's hidden labels

Some AI tools add a hidden label: OpenAI and Adobe add Content Credentials (C2PA) to the file, and Google adds an invisible watermark (SynthID). These help when they survive, but they are easy to lose: WhatsApp, Instagram and screenshots remove the file's labels, and many tools add none at all. A photo without a label is not proof that it is real.

4. Use a checker, and read the answer carefully

A detector looks at the pixels themselves, not at labels, so it still works on a forwarded or screenshotted copy. Real or AI gives one of three answers:

No detector is right every time, ours included. Use the answer together with where the photo came from, especially before you share it or act on it.

Signs that no longer work

More guides