Level: Beginner to Intermediate

Prerequisite: None

Ken Johnston

Vice President of Data, Analytics, and AI

Envorso

Most AI governance programs are working.  Well, working on paper, that is! The frameworks are documented, the principles are posted, the review boards are scheduled. And almost none of it touches the systems being shipped this quarter.

The disconnect is structural. Governance lives in slides; AI lives in production. Policies are written for the systems that existed when the policy was approved, not the copilots, agents, and pipelines that have shown up since. Reviews happen after deployment, when the cost of saying no is highest. Evidence, when a regulator or board asks for it, gets assembled from screenshots and Slack threads.

The result is a governance posture that satisfies no one. It slows the teams shipping the systems, fails to reassure the executives accountable for them, and produces nothing a regulator or auditor can verify under pressure.

This half-day workshop introduces AiGovOps — a way of running AI governance that works the way modern engineering works. Principles become controls. Reviews become automated gates. Opinions become evidence with a name and a timestamp on it. The operational stance is simple: if a control cannot be tested, owned, and audited, it is not a control.

Participants learn to apply the AiGovOps 30-day starter plan to a system they actually run. They identify the four owners every AI system needs (model, data, deployment, incident), map the controls that matter most for their highest-risk pillars, and design the first testable gate that produces audit-ready evidence by day 30. The exercises are deliberately small. The point is to leave with one working pattern, not a roadmap that lives in a binder.

The workshop draws on direct experience building AI programs at Microsoft and Ford, advising AI leaders across regulated industries, and operationalizing the AiGovOps framework as standards from NAIC, the EU AI Act, and the SEC have begun to bite. It reflects what works when governance is treated as engineering and what fails when leaders mistake documentation for control.

If your organization has policies it cannot operate, controls it cannot prove, and reviews that arrive after the systems have already shipped, this is where that changes.

You Will Learn

• Patterns of AI risk and how to assess patterns of HARM
• How to translate AI principles into testable, auditable controls owned by named people
• Why most AI governance programs collapse into "PDF theater," and the operational shifts that prevent this
• How to inventory and risk-tier your AI systems, including the shadow ones IT cannot see
• How to design governance gates that run as part of delivery rather than as a checkpoint after it
• How to produce the evidence regulators, auditors, and boards are beginning to demand without slowing teams down

Geared To

• Data and analytics leaders accountable for AI programs in production
• AI program owners and product leaders responsible for governed delivery
• Enterprise architects shaping AI platforms and operational patterns
• Governance, risk, and compliance professionals working with live AI systems
• Chief data officers, chief AI officers, and their direct reports
• Technical leads translating AI delivery into business and regulatory accountability
• Executive sponsors of AI initiatives who need governance evidence, not governance theater