Closed-door Roundtable: Detecting and Disrupting AI Agent Cyberattacks
Based on a forthcoming report with CSA, the roundtable will bring together government officials from multiple AISIs, cybersecurity industry practitioners, and AI security researchers to develop concrete approaches to adapting cyberdefences and information-sharing practices for a world with highly capable cyberagents.
Date: May 20th, 2026
Time: 3:00 PM – 5:00 PM SGT, followed by a networking reception
Venue: Sentosa Village Hotel, Singapore
Confirmed attendees: Staff from organisations including Interpol, SG CSA, MIT, Centre for the Governance of AI, Concordia AI, F5 Security.
Autonomous AI agents are increasingly able to perform offensive cyber operations. The recent announcement of Claude Mythos Preview and Project Glasswing by Anthropic demonstrates this as a clear and present concern. This roundtable aims to develop concrete approaches to detecting and disrupting AI-driven cyberattacks — from adapting cyberdefences to rethinking information-sharing practices for a world with highly capable cyberagents.
Due to limited venue capacity, only approved attendees will be able to attend.