When AI Acts
A one-day working session for leaders of small and mid-sized companies on governing the AI that now acts for your business.
Your AI tools have started to do things, not just answer questions: send, update, schedule, screen and approve. In California, “the AI did it” stopped being a defense on January 1, 2026.
Spend a day with other business leaders working out who decides, who can stop an agent, and what records you would need to show. Leave with a plan for the next 90 days.
Why spend a day on this now
AI now acts, not just answers
Tools that can send, change and approve things create exposure even when they break no rule. Most companies cannot yet see what their agents did, or stop them quickly.
California's rules have changed
AB 316 has applied since January 1, 2026. Rules on automated decision-making apply from January 1, 2027, and rules on automated decisions in hiring and employment already apply.
Governance sized for your company
You do not need a governance department. You need a few clear decisions, limits set outside the AI itself, and records you can produce when someone asks.
As of October 2026. The legal session is a plain-English briefing, not legal advice.
How the day runs
Timings may shift slightly before the day.
Registration and coffee
The question we answer by the end of the day
If your AI caused harm tomorrow, could you show what you reasonably did today to prevent it?
What changes when AI acts
How agents differ from chatbots, what a recent study of AI agents working together showed, and the five ways companies lose sight of what their agents do.
What changed on January 1
California's new rules, where liability lands, and what Colorado and federal action add.
Break
Your AI exposure map
Map the AI tools and agents your company runs, score each one for impact and autonomy, and pick the three to govern first.
Lunch
Tabletop exercise
At mixed tables, run a fictional mid-sized company through an afternoon in which its agents break no rule and still create legal, regulatory and reputational trouble. You play a role close to your own.
Break
Governance design
Design a minimum workable governance for a company your size: who can pause each agent, the limits set in the platform, the vendor terms to check, and a monthly review.
Scorecard and your 90-day plan
Score your company against eight questions and commit to actions with owners and dates.
Questions and close
You will leave with
- An exposure map of the AI your company runs, and the three uses to govern first.
- A plain-English view of California's 2026 and 2027 rules and what they ask of you.
- A minimum governance kit sized for your company: pause rights, platform limits, vendor terms to check, and a monthly review.
- Your score against eight questions every leadership team should be able to answer.
- A 90-day plan, and the workbook to run the exercise again with your own team.
Who should come
Owners, CEOs, COOs, CFOs, general counsel, HR and people leaders, and IT or security leads at small and mid-sized companies that use, or are about to use, AI that acts on their behalf.
No technical background is needed. Come with two or three colleagues if you can: the decisions cut across roles, and the day works best when the people who will act on the plan make it together.
Team rate: $345 per person for three or more people from the same company. Book your team’s places together.
Your facilitator
Neil W. Smith
Founder and principal, Decision Surface Group
Neil leads Decision Surface Group's work with executive teams on governing AI agents, using the firm's Three-System Model.
Questions
Is this legal advice?
No. The legal session is a plain-English briefing on what has changed. Take what you learn to your own counsel.
Do we need to be using AI agents already?
No. If your company uses AI tools that can send, update, decide or approve anything, the day applies to you. If you are about to start, it is the right time.
Is it relevant outside California?
Largely, yes. The legal session focuses on California and notes Colorado, federal action and the EU for companies that operate there. The exposure map, tabletop and governance design apply anywhere.
Is it technical?
No. The day is about decisions, controls and records, not code.
Will I have to share confidential information?
No. The tabletop uses a fictional company. You map your own company's AI in your workbook, and you choose what, if anything, to share at your table.
Does my ticket count toward a private workshop for my team?
Yes. Your ticket price is credited toward an in-house executive workshop for your leadership team booked by March 31, 2027.
How do team bookings work?
One person books all the places in a single checkout with the team ticket: three or more places at $345 each, on one card and one receipt. Then email neil@decisionsurface.com with the names and email addresses of the people attending.
Can my company pay by purchase order or invoice?
Yes. Email neil@decisionsurface.com with the names and email addresses of the people attending and your purchase order number, and we will send one invoice for all the places.
Can I transfer or cancel my place?
You can transfer your place to a colleague at any time. If you cancel by January 8, 2027, we refund 50% of the ticket price. After January 8, tickets are not refundable, but they can still be transferred.
Wednesday, January 13, 2027. Irvine.
Early bird $395 until December 18, then $495. Places are limited so that every table can work. Questions? Email neil@decisionsurface.com.