New and old methods to detect deepfakes before they enter the courtroom

01 Jul 2026
Theatre 2
In this lecture, we will discuss the usage of Artificial Intelligence for the creation and processing of digital media evidence. Illegal activities associated to this new technology are a real burden for the criminal justice system. You will learn how deepfakes are created and understand the concepts beyond Generative Adversarial Networks (GAN’s) and "text-to-image" stable diffusion. You will learn how the provenance of the evidence plays a pivotal role in determining its authenticity. We will also discuss how neural networks  can be trained to spot deepfakes and how we can still use older authentication methods, such as shadows, reflections and perspective analysis, to find evidence of AI manipulation.
Speakers
Alexander Fehrmann
Alexander Fehrmann, Trainer and Technical Support Specialist - Amped Software

Sponsored by:

Amped