2026-05-09 –, Grégr Stage
As AI-generated content becomes indistinguishable from reality, society must evolve beyond detection strategies. This talk explores how we can build resilient information systems and social institutions for a future where synthetic media is ubiquitous. From reimagining trust frameworks to developing new verification paradigms, we'll examine what societal transformation are needed for us to thrive.
We stand at a critical inflection point in human information history. As generative AI rapidly approaches - and in some domains, achieves human-indistinguishable output across text, images, audio, and video, the fundamental assumptions underlying our information ecosystems are being challenged. The question is no longer whether AI-generated content will be detectable, but how society will function when it isn't.
This talk examines the broader societal implications of living in a world where synthetic and authentic content are indistinguishable. Beyond the immediate concerns around misinformation and deepfakes, I explore how critical sectors - from satellite imagery and scientific evidence to journalism and legal documentation - face unprecedented vulnerabilities. These challenges extend beyond technical detection problems to fundamental questions about trust, verification, and truth in the digital age.
The talk concludes by exploring what a resilient information society looks like: one that acknowledges synthetic media as permanent, develops adaptive verification systems, strengthens institutional safeguards, and creates policy frameworks that balance innovation with protection. This is not about preventing the future, but about building the social, technical, and regulatory foundations that allow society to thrive within it. The path forward requires coordinated action across technology development, education, policy, and cultural adaptation - a transformation as significant as previous shifts in human communication.
Giulio Corsi is a Research Associate at the Leverhulme Centre for the Future of Intelligence at the University of Cambridge. His research focuses on quantitatively monitoring and mitigating risks from AI to information ecosystems.