# AI operating model

*When the model changes, what remains?*

Ownership, governance, skills and routines that make AI dependable, whichever model you run.

Outcome: Build adaptability. Delivered by OneNine Advisory. Platform: any.
People: [jesper-andersen](/people/jesper-andersen), [aksel-watson](/people/aksel-watson), [ferdinand-drobik](/people/ferdinand-drobik).
Cases: [medtech-ceo-workshop-human-plus-ai](/cases/medtech-ceo-workshop-human-plus-ai).
Drawn answer: Model (answers) → Agent (acts) → Owner (a named human decision).

The model can change. Knowledge, skills, control and accountability still need an owner. We help leadership teams decide what AI may do, who checks the results, where it runs (cloud, private cloud, on-premises) and how it is governed under GDPR and the EU AI Act.

We run this as a CEO workshop first, then a 90-day path that starts with one bounded process.

Canonical: https://oneninenordic.com/services/ai-operating-model
Source: OneNine content model, updated 2026-11-04