AI That Holds Up: Making AI work in a regulated business
Taking an AI use case from idea to defensible, under regulation.
About this session
Most AI use cases in regulated organisations don't fail in the model. They fail in the questions nobody asked early enough. What data went in? What does the vendor's contract actually allow? Who checked the output? And what could you show an examiner, an auditor or your board six months later?
This is a working lunch for the people who have to answer those questions, co-hosted by Dr. Dominik Kalisch and Brian Dietrich. It is about making AI work in a business where secure and compliant aren't optional: how to pick the use case, what to settle before the pilot, and what to keep so the decision still holds up later.
It is ninety minutes around one table with nineteen peers from regulated industries. There's a short talk, then a case clinic built on the use cases attendees bring. Lunch is provided. Nobody is selling anything in the room, which is one reason seats are by application.
What you'll leave with
- Five questions every AI use case has to answer before it goes live: data, vendor, oversight, evidence, and owner and exit
- A one-page use-case check you can run on your own idea this week, printed for you on the day
- The questions to put to any AI vendor before you sign, on retention, training on your data, where data is stored, and subprocessors
- Where regulated organisations most often get caught: staff using personal AI tools, vendor terms nobody read, and decisions nobody wrote down
- A table of regional decision-makers who answer to the same kinds of examiners and auditors
Who it's for
- Executives, owners and senior leaders who decide on, sponsor or approve AI in an organisation that is regulated, examined, audited or accredited
- Heads of operations, risk, compliance and IT who will be asked to sign off on it
- Financial services, healthcare, education, and any other field where someone outside the building will one day ask "show me how you decided this"
- Lubbock and West Texas organisations, whether AI is in use already or only under discussion
Agenda
- 12:00 Lunch and introductions around the table One line each: who you are, and the use case you brought.
- 12:15 Welcome
- 12:20 Why AI use cases fail The three things every working AI project shares, in ten minutes.
- 12:30 Where regulated organisations get caught Shadow AI, vendor terms, missing records, and accountability that sits with nobody.
- 12:45 Five questions every use case must answer Data, vendor, oversight, evidence, and owner and exit: what each one asks, and what a good answer looks like.
- 13:00 Case clinic Run your own use case through the one-page check with the person next to you. Then we work through one or two from the room, together.
- 13:20 Questions and close
- 13:30 End Stay and talk if you like.
All times CT.
Who's presenting
-
Dr. Dominik Kalisch Co-Founder, Alephnet · Faculty, Rawls College of Business at Texas Tech · former Chief Analytics Officer, NORD/LB
Read full bio
-
Brian Dietrich President, Sentinel Tech Consulting and Training LLC
Questions people ask
Why is it by application?
Because the session works best when everyone at the table carries the same kind of responsibility. We keep it to decision-makers in regulated organisations, at most two per organisation, and we don't seat vendors. Applying takes about two minutes. We reply within two working days.
Who should apply?
Anyone who decides on, sponsors or signs off on AI in an organisation that answers to a regulator, an examiner, an auditor or an accreditor. That includes banks and credit unions, insurers, healthcare providers, schools and universities, and others. If you're not sure your field counts, apply and tell us who regulates you.
I sell AI or technology services. Can I come?
Not to this one. Apart from the two hosts, everyone at the table is there as a peer. Keeping vendors and consultants out is what lets people speak openly about what they're actually doing. We hope you understand.
What does it cost?
Nothing. Lunch is on us, and there's nothing to buy in the room.
Is this legal or compliance advice?
No. We'll talk about how regulated organisations approach AI and point to the rules that commonly apply, such as model-risk guidance for banks, HIPAA for health data, FERPA for student records, and Texas's own AI legislation. How they apply to you is a question for your counsel and your compliance team. We'll help you arrive with better questions.
What happens to the use case I describe in my application?
We use it to plan the session, and with your agreement as an anonymous example on the day. It stays in our CRM with your application, it isn't shared with other attendees, and we never publish it. Please leave out anything confidential. A sentence or two is enough.
Can I bring a colleague?
Yes, up to two people per organisation. Each applies separately so we have their details and any dietary needs.
Is this a sales pitch?
No. Dominik and Brian will each say what Alephnet and Sentinel Tech do once, at the end. If you'd like to talk afterwards, they'll both be around. If not, you won't hear from us beyond a short follow-up with the materials.