SI square

Creating science-based standards for AI-based opinion and agency manipulation technologies

, -

the fab, 4th floor of Petal 2, Maison de la Paix, Geneva

The session will start by discussing studies on agency from neuroscience and psychology including findings of innate needs for agency as well as providing a brief overview of Bayesian theories of agency. You will learn an overview of the neural correlates of voluntary action and highlight the increasing risk of manipulation from powerful human-behavior prediction AI models that use social, biometric as well as neural datasets.  You will also get some suggestions including formal approaches for defining agency via causal modeling research and quantifying the manipulative and agency-engineering components of AI technologies including social media and knowledge recommenders.

Add to Calendar



Catalin Mitelut, immigration lawyer and computational neuroscientist, will advocate for the need for (international) standards for the protection of human agency from AI-based agency and opinion manipulation technologies.