Could you run AI on your own terms?

Your score comes only from your own answers and the rules below. It is not a benchmark against other companies, and we have no data on how others score.

1. Every question adds points to one dimension. Single choice earns the listed points (0 to 4). Each ticked option of a multiple-choice question earns 1 point. A scale answer earns its number minus 1.
2. A dimension score is the points you earned divided by the points available in that dimension, as a percentage.
3. The overall score is all points earned divided by all points available. Every dimension has the same number of questions, so each counts equally.
4. Levels follow the overall score: Starting 0 to 29, Developing 30 to 54, Established 55 to 79, Leading 80 to 100.
5. Recommendations come from the rules listed per check. The weakest dimensions rank first, and the top three are shown.

Points per question

Each answer shows the points it earns on the dimension of the question.

1. 01 / Data ownership Who owns the data you want AI to learn from or answer with? Up to 4 points

   • It is unclear 0 pts
   • We own some of it, third parties own the rest 1 pt
   • We own it, but our rights are not documented 3 pts
   • We own it, and rights and purposes are documented 4 pts

2. 02 / Data ownership Which controls exist for that data? Up to 4 points

   • Classification by sensitivity 1 pt
   • A named owner for each dataset 1 pt
   • Retention periods 1 pt
   • Masking or pseudonymisation of personal data 1 pt
   • None of these 0 pts

3. 03 / Residency How clear are your residency requirements? Up to 4 points

   • We have not looked into it 0 pts
   • We assume EU only, but have not checked 1 pt
   • They are written down per dataset 3 pts
   • They are written down and checked against supplier contracts 4 pts

4. 04 / Residency Which hosting options could you accept for an AI workload? Up to 3 points

   • An EU region of a public cloud, such as Azure 1 pt
   • A European AI cloud region, for example Nebius 1 pt
   • Our own servers or data centre 1 pt
   • None of these 0 pts

5. 05 / Model control How do you choose AI models today? Up to 4 points

   • Whatever our tools include 0 pts
   • One hosted vendor 1 pt
   • We compare several and can switch 3 pts
   • We compare several and can run open-weight models ourselves 4 pts

6. 06 / Model control How well can your team judge whether a model is good enough for a task? Up to 4 points

   Scale 1 to 5 earns the answer minus 1 (0 to 4 points).

7. 07 / Infrastructure Who would run the models and GPUs? Up to 4 points

   • Nobody yet 0 pts
   • IT would learn it 1 pt
   • A partner runs it under our contract 3 pts
   • In-house or a partner, with a run book and on-call cover 4 pts

8. 08 / Infrastructure How ready is your platform for a private deployment? Up to 4 points

   Scale 1 to 5 earns the answer minus 1 (0 to 4 points).

Recommendation rules

• Data ownership scores below 50 → recommend: Document your data rights and owners
• Residency scores below 50 → recommend: Map residency requirements to hosting options
• Model control scores below 50 → recommend: Build an evaluation set before you pick a model
• Infrastructure scores below 50 → recommend: Prepare a private foundation
• Overall score is 70 or higher → recommend: Choose the hosting and run a self-hosted pilot

Levels

• Starting (0 to 29): Key questions about rights, location and hosting are still open. Answer those first. If no rule forces self-hosting, a hosted model may be the faster start.
• Developing (30 to 54): Some of the groundwork is done. Writing down residency and data rights lets you choose between Nebius in a European region, your own servers or Azure in an EU region.
• Established (55 to 79): You can choose a hosting option and judge models with evidence. A self-hosted pilot on open-weight models is realistic.
• Leading (80 to 100): Rights, residency, evaluation and a private foundation are in place. The remaining work is choosing the hosting and model and running them with a clear handover.