October 17 lootings: national police spark controversy with an AI-retouched photo of a seizure

On October 17 and 18, approximately 3,000 businesses were affected, and many were looted. Vinted and Leboncoin are flooded with listings for stolen goods, and fake fundraising campaigns supporting merchants are appearing. Amid all the chaos, the police are tracking down and arresting looters and handlers of stolen goods. In Val-d’Oise, the National Police arrested 12 of them and displayed the products put up for resale.Except that the published image of the products was generated by artificial intelligence, even though the products do exist.

The National Police publish an AI-generated image: internet users notice this and fuel the controversy

It was on X on October 20 that the Val d’Oise National Police account published an image. It shows clothing and video games on display, the result of the arrest of 12 resellers in the department. However, some details do not add up: the brand logos are full of flaws, and the lettering on the cover of Elden Ring Nightreign is rather strange.

#ValDOise | Present. Mobilized. Committed. This weekend, @PoliceNat95 alongside @prefpolice after the looting in #Paris:

👉 1️⃣2️⃣ resellers arrested in the 95

📦 Seized: ⌚ high-value watches, 🧢 designer caps, 👓 #LouisVuitton and other luxury products still fitted with anti-theft tags https://t.co/BODLxxVPoT

— National Police 95 (@PoliceNat95) October 20, 2026

By running the image through OpenAI’s or Google DeepMind’s verification tool, one notices that it was generated using OpenAI’s tools. OpenAI itself says so. Indeed, a SyntID watermark from OpenAI appears on this photo: the photo has indeed been edited or generated with ChatGPT. For the past few days, it has been possible to use these SynthID tools for free. For reference, SynthID is a technology developed by Google DeepMind that adds invisible digital watermarks to AI-generated content.This makes it possible to detect them later, particularly to debunk fake news. SynthID works on audio, video and images, as is the case here.

The police deny it and give the real explanation

It took nothing less for hundreds of internet users to be outraged on X. All of this in a context where part of the population distrusts the police, which is accused of violence and corporatism. Indeed, many people under the first tweet denounce manipulation by the police, which would fabricate false evidence to justify the arrests of protesters and vandals or looters.

The next day, the official National Police account published a denial: “This photo is not a 100% AI creation.” The account explains that its Val d’Oise branch modified an image using an AI app “to crop it and remove the superfluous elements (chairs and air conditioner).” All without posting a reply under the first tweet. But the damage is done: the original post surpassed one million views in less than 24 hours, while the denial post has only about a quarter as many views.

👉🏻 Follow tech news in real time: add 01net to your sources on Google, and subscribe to our WhatsApp channel.