Make agentic AI governable, auditable and defensible.
Your AI agents now plan, delegate and act. We help organizations keep human authority over what those agents do, and keep the evidence that proves it.
If your AI agents cause harm tomorrow, can you show what you reasonably did today to prevent it?
Most leadership teams cannot yet. The gap is not what the agents can do. It is what the organization can see, control and prove.
The accountability gap is widening.
Agents act at machine speed, across your systems and alongside other agents. Your ability to understand, control and defend what they do does not keep pace unless you build it. We call this asymmetric illegibility, and it takes five forms.
Detection is not containment.
In a company-run simulation published in September 2026, no society of AI agents withstood all three staged attacks. Agents that recognized an attack still stored it, shared it or acted on it.
- Operational
- You cannot reconstruct what the agents did.
- Causal
- You cannot say which decisions produced the outcome.
- Intentional
- You cannot tell whether the agents still serve the objective you set.
- Communicative
- You cannot follow what agents tell each other, even with every message logged.
- Governance
- You cannot show which controls applied, or whether anyone could have stepped in.
“The AI did it” is no longer a defense.
Since January 1, 2026, California law has barred defendants who developed, modified or used AI from arguing that the AI caused the harm on its own. Causation and foreseeability can still be argued, so evidence of what you knew and what you did decides the case.
California AB 316 takes effect (Civil Code §1714.46).
California's rules on automated decision-making technology apply, and Colorado's replacement AI law takes effect.
EU AI Act obligations for high-risk systems listed in Annex III apply, as amended in 2026.
As of October 2026. A summary, not legal advice: confirm how these rules apply to you with your counsel. The full calendar, with sources.
Wednesday, January 13, 2027, in Irvine, California
When AI Acts
A one-day working session for leaders of small and mid-sized companies. Map the AI that acts for your business, test your decisions in a live exercise, and leave with a 90-day plan.
Early bird $395 until December 18.
One model for intent, agency and governance.
Most AI programs blur three different jobs: deciding what the organization wants, letting agents do the work, and governing the result. The Three-System Model separates them, connects them and makes each one testable.
Human intent
What you want the agents to achieve, the limits they work within, and who has authority to decide.
Machine agency
How agents plan, delegate and act, through controlled delegation and structured hand-offs.
Defensible governance
- Prevent what you can
- Intervene when you must
- Substantiate what you did
Start with a half-day. Build from there.
Each engagement delivers part of the model. Most clients begin with their leadership team in one room.
-
Executive workshop
Map the agents you run, test who decides under pressure, and leave with a 90-day plan.
-
Agentic Safety Audit
An inventory of every agent and AI workflow, scored for exposure across all three systems, with a gap report.
-
Trust Harness
Intent specifications, delegation limits, pause rights, and Continuous Inquiry with measured detection rates.
-
Evidence Fabric
Decision records built from what agents did, to a standard an examiner, auditor or court can follow.
Governance that lets you do more, not less.
Control you can test
Limits enforced outside the model, the power to pause, and oversight whose catch rate you have measured.
Evidence that stands up
Records built from what agents did, ready for an examiner, an auditor or a court.
Confidence to expand
Clear boundaries let you give agents more autonomy where it pays.
Built for the leaders who sign the filings, and the teams who run the agents.
CEOs, CFOs, COOs, general counsel, chief compliance and risk officers, CIOs and security leads.
Our in-house work focuses on regulated sectors, including banking and lending, insurance, healthcare and wealth management, from mid-sized firms to larger institutions. Our public sessions are open to any company putting AI agents to work.
We are not a software vendor, a law firm or a general AI strategy shop. We work alongside your counsel and your technology teams, with whatever agents you run.
Before autonomy outruns accountability.
Book a half-day with your leadership team, or tell us where your agents stand today.