For the first time in Mayenne, the West Data Festival held the "Trial of AI", an immersive and realistic enactment aimed at exploring the risks, potential pitfalls, and solutions in response to the rapid development of this technology. Rigorously adhering to the codes of a real court, the event placed artificial intelligence on the defendant's bench for generating a deepfake compromising a political figure.
In turn, lawyers, scientists, and experts took the stand to defend or accuse AI. The highlight of this original experience was the presence of artificial intelligence itself in the form of a virtual avatar, directly responding to the questions from the judge and assessors.
We invite you to watch the full replay of this exceptional session of the West Data Festival, now available online on YouTube, offering everyone the unique opportunity to relive this exciting and unprecedented debate about the ethical and societal challenges of AI.
Translated from Replay West Data Festival : un procès de l'intelligence artificielle aux enjeux bien réels
To better understand
What is a deepfake and how are they technically generated?
Deepfakes are synthetic media created by AI algorithms, mainly generative adversarial networks (GANs). These models learn from existing data to generate new images or videos that closely resemble reality.
What is the impact of European regulations on the use of deepfakes?
The European Union, through its Digital Services Act, imposes obligations on online platforms to label and detect deepfakes, aiming to protect users from misinformation and privacy violations.