1. Skip to content
  2. Skip to main menu
  3. Skip to more DW sites
TechnologyGlobal issues

Detecting deepfakes with chips and AI

04:11

This browser does not support the video element.

July 16, 2026

Most deepfakes can still be identified by an experienced eye. But how long will that remain the case? New technologies are now aiming to verify authentic images and reliably detect digital forgeries.

A team in Basel is developing a camera sensor that generates an invisible authenticity "watermark” with every image captured. This makes it possible to distinguish genuine photos from AI-generated fakes. At the same time, PXL Vision uses AI models trained to identify typical patterns found in synthetic images, from pixel structures to subtle color transitions. The technologies are designed to protect social media platforms, banks, and telecom providers from fraud. As deepfake methods continue to evolve, experts must constantly retrain detection systems to keep pace.

Skip next section DW's Top Story

DW's Top Story

Skip next section More stories from DW

More stories from DW