As generative models continue to improve, deepfakes are becoming increasingly realistic and harder to distinguish from authentic content. This guest talk by the AI Service Center Berlin-Brandenburg explores how deepfakes are created and how AI-based techniques can be used to detect them.
Jonas Bernhard introduces Clyravision, a multi-agent system that combines different models and methods to identify deepfakes in an interpretable and transparent way. The talk is aimed at anyone interested in trustworthy AI, media integrity, and the challenges posed by generative technologies.
Event Details
-
Title: Guest Talk: Automated Deepfake Detection with AI
-
Format: In-person guest talk
-
Date: Tuesday, March 17, 2026
-
Time: 5:00–6:30 PM (CET)
-
Location: Hasso-Plattner-Institut
Campus Kitchen One, Building L, Room L-1.02, 14482 Potsdam -
Organizer: AI Service Center Berlin-Brandenburg @ HPI
-
Duration: 1 hour 30 minutes
-
Admission: Free, registration required
-
Note: Photo, audio and video recordings will take place during the event
Hasso-Plattner-Institut
Campus Kitchen One, Building L, Room L-1.02, 14482 Potsdam