Resources

Who is liable when an AI agent goes rogue?

Rogue AI agents are operating in your organisation right now. In this panel discussion from the Okta AI Summit 2026, legal and cyber industry experts Chang Lim (Mallesons) and Eddie Smith (Versent) speak to Okta's Stephanie Barnett to expose the operational, legal, and identity risks you can't ignore.
Autonomous AI demands a completely different security playbook. Over-privileged access. Uncontrolled decision-making. Agent drift. Traditional controls don't cut it.  The uncomfortable truth is your board is liable for every action your AI takes—full stop.

Discover how to secure AI adoption with the identity controls that matter: exact privilege, mission-based access, complete auditability, and immediate kill-switches.

Watch the session and learn what robust identity architecture actually looks like in an agentic AI world.

Topics

  • Video
  • ai agents
  • Compliance
  • Cybersecurity
  • AI Agents
  • Non-Human Identities
  • Threat Intelligence
  • Threat Detection