18 June: From Botnets to AI Swarms: Tracking Emerging Risks Through an AI Influence Observatory

The combination of large language models and AI agents could add a new layer of risk to the threats AI already poses to the information environment, potentially reaching levels we are only beginning to imagine. Drawing on recent research on malicious AI swarms, this webinar opens a discussion on how AI-enabled disinformation may evolve and scale in sophistication and coordination: from persistent AI personas and AI-supported botnets to cyborg propaganda through verified human accounts, adaptive engagement testing, and campaigns designed to manufacture the appearance of public consensus. But the session does not stop at mapping the risks. It also aims to open a timely debate on possible responses, and on the evidence, infrastructure, governance, and safeguards needed, with the AI Influence Observatory as a flagship proposal to make these emerging risks more visible, comparable, and democratically accountable. The discussion explores how researchers, civil society, journalists, platforms, and public institutions can jointly study and respond to this emerging risk, including through an operational taxonomy, a curated repository of cases, simulation environments to study swarm behaviour, behavioural and coordination-based detection signals, and prototype “AI Shields”. Rewatch our discussion! Speaker: Daniel Thilo Schroeder, Researcher at SINTEF Moderator: Raquel Miguel Serrano, Senior Researcher, EU DisinfoLab The opinions expressed are those of the speakers/authors and do not necessarily reflect the position of EU DisinfoLab. This webinar does not represent an endorsement by EU DisinfoLab of any organisation.