OC Tech Week 2026 · Series I / SYSTEMS · Registration Open
When AI Goes Rogue: Protecting Your Clients and Your Team When Algorithms Take Control
When AI can read information, trigger workflows, and act on a team's behalf, practical oversight matters. Join Tory Erickson for an Executive AI War Game and interactive roundtable exploring information boundaries, permissions, spending controls, human review, vendor responsibilities, and incident response.
Understand the system. Control the operation. Prepare for change.
Ticketing and registration are handled through Luma.
At a glance
- Wednesday, October 21, 2026
- 11:00 a.m.–1:00 p.m. Pacific; program begins at 11:30 a.m.
- UCI Beall Applied Innovation @ the Cove, 5270 California Ave #100, Irvine, CA 92617
- Interactive Executive AI War Game, briefing, and roundtable workshop; event capacity 40
- Catered lunch and refreshments included
- Registration: $75 minimum; $195 suggested contribution. Check Luma for current availability and ticket terms.
What participants examine
The briefing moves through six areas where AI systems most commonly exceed their intended boundaries — and where organizations are most exposed when they do. Each area is examined through a practical lens: what the risk looks like, what controls exist, and what questions executives should be asking their teams.
Agent permissions
What actions can an AI agent take on behalf of your organization, and who authorized them? Participants examine how permission scopes are defined, granted, and revoked — and what happens when they are not.
Sensitive-data handling
AI workflows often touch data that employees would never share externally. This segment examines how sensitive information enters, moves through, and exits AI pipelines — and the controls that should govern each step.
Workflow loops
Autonomous agents can enter loops that consume resources, repeat actions, or escalate without human awareness. Participants explore how to detect, interrupt, and prevent runaway workflows.
Vendor responsibilities
When an AI vendor's model or infrastructure fails, who bears the consequence? This segment maps contractual and practical responsibility across the vendor, platform, and deploying organization.
Output review
AI-generated outputs — text, decisions, code, communications — require human review before they carry organizational weight. Participants examine where review is required, who performs it, and what "review" actually means in practice.
Incident response
When an AI system behaves unexpectedly, organizations need a response plan. This segment covers detection, containment, documentation, and the decision of when to involve legal counsel.
Continue with your own AI workflow
Use the Executive AI Control Scorecard to review a defined AI use case. Companies, teams, and individuals can also request a private, facilitated scorecard discussion. Scope, timing, deliverables, and fees are agreed separately.
This briefing is general executive education. It does not constitute legal advice, carry MCLE or CLE credit, or create an attorney-client relationship. Viewing this page or attending the briefing does not establish a legal engagement with T. Stanford Law. Attending does not guarantee liability protection or establish compliance with any standard. Professional Disclosures