The video series

Frode Stenstrøm and Adrian Sjøholt compared tools from Google, OpenAI and Anthropic through tasks in project planning and documentation. Examples included requirements documents, milestone plans and reporting based on a shared assignment.

What this means for the organisation

The results show how particular model versions handled the selected instructions at the time of recording. Project managers must check assumptions, dependencies and numbers against the actual project basis and retain traceability when editing an AI draft.

The series used the same project basis for requirements, milestones, WBS, responsibility matrix, risk and presentation tasks across models. This makes the comparison concrete but also tied to the selected versions and prompts. Project dependencies and decisions still need traceability to people and approved documents.

NIST AI RMF structures risk management around govern, map, measure and manage. It provides a vendor-independent frame for reading ProsessPilotene’s guidance and model tests. The demonstrations show particular tasks; the framework also requires organisations to assess context, consequences, measurement and responsibility throughout the lifecycle.

A pilot needs a bounded dataset, expected answers, difficult counterexamples and a named decision owner. Errors should be classified by consequence rather than merely counted. Model and prompt versions should be retained so the organisation can explain changed results before giving an agent wider scope to act.

Sources

Primary source: ProsessPilotene, “Slik bruker du AI i prosjektarbeidet,” 1 September 2025 (Norwegian).

NIST, AI Risk Management Framework 1.0, 26 January 2023.

Join the discussion

Which experience from your own organisation is most relevant to this story? Share what you learned.