How to tell if that Elon-Zuck-and-Bezos-at-a-bar photo is a fake

Based on this article by Matt Novak posted on Gizmodo.

Amidst a surge of highly realistic, yet fake, viral photos, such as fabricated images of tech CEOs gathering in mundane settings, Google has given the public a powerful new detection tool. The company has unlocked its "SynthID" watermark detection within Gemini, allowing anyone to upload an image and instantly check if it was generated by Google's AI tools.

This invisible, pixel-embedded watermark provides a trustworthy method for the average user to verify the authenticity of an image, bypassing the failures of current, unreliable AI detectors. Though this only addresses content created by Google's generators, it offers a crucial defense line against the increasingly sophisticated deepfakes flooding social media platforms.

Check out this article.

Related Stories