Chapter 01
Research agenda
Programs cover perception spoofing, training-data poisoning, LLM jailbreaks, and autonomous response agents that stay inside policy. Publications include our adversarial robustness research and ongoing red-team reports.
Laboratory
Security Lab
Adversarial ML and autonomous defense.
Red-team
Culture
Cross-lab
Hardening
Product
Transfer
Chapter 02
Instruments
Attack libraries, physical test stages, isolated evaluation clusters, and shared threat ontologies used by product security reviews before launch.
Chapter 03
Engagement model
Enterprise and defence partners can commission targeted red-teams. Findings, when permissible, flow back into Cyber Intelligence detection content and AI Security mitigations.
Shared instrument set
Six laboratories. One research fabric.
Security Lab shares compute, evaluation, and publication tooling with every other Zansoc lab — so discoveries compound across the stack.
Visit all labs