Hear three key risks every IT leader must consider when governing AI for their organization. data security, unintended actions from non‑deterministic agents, and attacker manipulation via prompt injection/data poisoning. You’ll get a clear, actionable framework from Box CTO Ben Kus and Senior Product Marketing Manger Meena Ganesh, plus guardrails to make agents operationally safe: secure access (authorization/permissions, least privilege, auditing), restrict dangerous tools, and require human‑in‑the‑loop for high‑risk actions. Watch the full conversation at the link in the comments