Are you still asking, "Can we jailbreak the model?" when the real vulnerabilities are already inside your pipeline? Security teams keep treating AI risk like a chatbot parlor trick instead of identifying the gap where real breaches happen. Attendees will learn which "obvious" AI security questions are actually wasting their time, and which unglamorous ones deserve real budget and attention. You'll walk away with a sharper checklist for what actually matters when securing AI systems in production. Come find out if your team's threat model is protecting you or just making you feel safe.
*Sponsored by Gray Swan
Virtual
As AI speeds up deopment, teams can add dependencies faster than they can review them. Attackers exploit that gap with malicious packages disguised as familiar components, plugins, and developer tools. Sonatype logged 149,329 malicious packages in Q3 2026, about 1,600 a day, bringing the total tracked since 2017 to nearly 2 million. Among packages with overtly malicious behavior, nearly 75% involved droppers, secrets theft, or both. A separate analysis of 4,309 packages found attackers using believable, ecosystem-adjacent names to evade notice.
In this session, we’ll examine how those attack chains unfold, including a malicious PyPI package published by an AI agent and the Mini Shai-Hulud campaign’s use of stolen npm credentials to republish compromised packages. You’ll learn practical ways to prevent malicious components from executing, detect suspicious behavior, limit credential exposure, and investigate what happened after a package entered your environment.
*Sponsored by Sonatype
Virtual
