As AI systems become more sophisticated, they also become attractive targets for malicious actors. SecureAI researches defense mechanisms — against model poisoning, adversarial attacks, data inference, and model extraction — that protect AI throughout its entire lifecycle.
CAIRLAB Compute Share InitiativeWant to get involved as a student, researcher, or engineer?
Apply to Get Involved → Back to Projects SecureAI
Building Trustworthy and Resilient AI Systems
Adversarial ML DefenseRobust Model DevelopmentVulnerability AnalysisSecure Deployment