Photo of elderly French couple ‘stripped and robbed’ during riots is AI

Several of our readers have messaged us about a picture of an elderly couple in their underwear that has been shared thousands of times on social media with claims they were stripped and robbed during the recent riots in France.

This picture isn’t real and was made with artificial intelligence.

The image has been shared with captions claiming the couple were “stripped of everything, right in the middle of the street” and accusing “far left and migrant rioters” of stripping them.

There have been mass protests across France since late September, led initially by students complaining about poor classroom and learning conditions. On 17 October, a national day of action backed by student groups, unions and the Yellow Vests movement led to a number of violent incidents.

But there are no official reports confirming such an attack on an elderly couple like this. The public prosecutor's office told French newspaper Libération on 19 October that it had (translated with Google) “received no reports to that effect to date”.

How do we know it was made with AI?

When we ran both variations of the image through Google’s SynthID detector, which allows anyone logged in to check if an image, video or audio file was made or edited with AI using Google, OpenAI, NVIDIA or Kakao products, they flagged as being made or edited using Google’s AI tools.

There were several visual clues too, leading us to conclude the original was AI-generated. The elderly woman appears to have three fingers plus a large, deformed looking thumb. The masked person on the left also seems to have two heads but only one body.

The earliest version of the image we found online was posted on X 17 October as part of a montage that was flagged on the platform as having been “made with AI”. The white border remains in all of the posts we’ve seen sharing the image online, strongly suggesting this is where it originated from.

Before sharing content like this that you come across, it’s important to consider whether it comes from a trustworthy and verifiable source. Our misinformation toolkit contains tips on how to do this.