Learn about the first documented attack campaign targeting AI workloads through a 38-minute LASCON conference talk that reveals the active exploitation of CVE-2023-48022 in the Ray AI framework. Explore how attackers have taken advantage of this disputed vulnerability over a seven-month period, impacting various sectors including education, cryptocurrency, and biopharma. Gain insights into the attack techniques, compromised sensitive data, research methodology, and real-world case studies that demonstrate the security implications of rapid AI technology adoption using open-source frameworks. Understand the critical lessons learned from this campaign and why many development teams and static scanning tools failed to detect this significant security threat.