Why was change necessary?

In autumn 2025, Suomen Ekonomit was looking to shift its own AI journey into a higher gear. Leadership had woken up to how fast working life is changing — and to the fact that advancing the change would require concrete decisions in their own organisation as well.

Ekonomit already had several technical AI experiments behind them, along with ethical guidelines and instructions. It was precisely these experiments that had sparked the understanding that technical work alone is not enough — organisation-level policies, responsibilities and a clear governance model are needed.

---

What goals were set for the project?

Head of IT Krista Karusalmi was tasked with executing the project plan approved by the leadership team. The project’s goal was to launch a more structured AI journey, so that AI could be integrated as a natural part of Ekonomit’s everyday work and strategy.

In the big picture, the aim was control and clarity in a way that enables and inspires — not slows down.

---

How did we approach the whole?

Astu Labs brought an outside perspective, expertise and a structured approach to the work. The work was divided into several phases:

Phase 1 — Current-state analysis of AI readiness

Astu Labs carried out a current-state analysis of Ekonomit’s AI readiness, which laid the foundation for planning and prioritising the actions to come. The analysis surfaced both strengths and areas for development, on which the actual policies were then built.

→ See an example of an AI maturity assessment by Astu Labs

Phase 2 — Building the policies and the governance model

Based on the current-state analysis, the policies and boundaries for AI use were clarified, version 1.0 of the governance model was built, and an AI coordination group was formed.

Phase 3 — Defining the roadmap

As the final output, a roadmap was built for a considered, careful rollout of the change. The roadmap serves as a practical tool for the leadership team and the coordination group in support of prioritisation. It was planned for roughly six months ahead and included recommended actions in four areas: 1) developing staff skills, 2) refining the governance model and developing processes, 3) leading the change, and 4) growing technology and data capabilities.

---

What did the project deliver?

The project’s key outputs were:

• clear AI policies and boundaries for everyday use
• governance model version 1.0
• a named AI coordination group
• a roadmap for carrying out the organisation-level change

The final output was presented to Suomen Ekonomit’s leadership team in November 2025, on the agreed schedule. In total, the engagement took about two months.

---

The client’s voice

“I knew from the start that we needed someone on the journey who could lead us in the right direction. The challenge was finding a partner in the wild mass of providers — one who would bring us real added value. At first, Astu seemed like a small player, and they didn’t have a huge number of references, which added to the doubt. We decided to start the collaboration anyway, because Astu managed to stand out with their professional conversations and the material they delivered — their views were surprisingly well structured and inspired confidence.”

“Once our shared journey began, it quickly became clear that starting the collaboration was a smart decision from Ekonomit. Astu brought us exactly the kind of holistic, calmly structured outside perspective we needed. They helped us articulate and crystallise the things we had already realised in our own earlier experiments, but hadn’t yet shaped into organisation-level policies or governance models. On top of that, they had the ability to bring concreteness and structure — not just visions.”

“I was extremely happy with the final output — it was clear, well grounded and structured exactly the way our organisation needed. Working with Astu was effortless and expert, and the dialogue stayed open throughout the journey. I can recommend Astu for similar engagements, especially when you need controlled, well-structured support for your organisation’s AI roadmap and governance model.”

— Krista Karusalmi, Head of IT, Suomen Ekonomit

---

What did we learn?

The project crystallised several lessons that apply more broadly to building an AI governance model:

• A governance model doesn’t slow things down — it enables and inspires, when it’s built properly the first time
• Technical readiness and leadership readiness advance at different speeds, and the gap grows every month unless it’s deliberately closed
• The lack of shared policies blocks the move from experiments to organisation-level use — even when ethical guidelines already exist
• An outside perspective helps put into words what is already intuitively known, and turn it into organisation-level policies

---

Who is this case useful for?

This case is especially relevant for:

• Expert organisations and associations where knowledge work is the core of the operation
• Leaders and supervisors wondering how to make AI a governed part of operations
• Organisations that already have AI experiments and ethical guidelines, but lack an organisation-level governance model and a shared direction

---

How does a project like this usually proceed?

The Suomen Ekonomit case reflects a typical path when building an AI governance model:

1. Mapping the current state and readiness — what AI is already in use, where to start from
2. Clarifying the goals — why the governance model is being built, what it should achieve
3. Defining shared policies and responsibilities — who decides what, and within which boundaries
4. Building the governance model and the coordination group — a documented model plus a named body with a mandate
5. Building the roadmap — prioritised actions for a considered rollout
6. The leadership team’s decision on next steps — the governance model gains official status in the organisation

---

Considering something similar?

If your organisation is experimenting with AI but the whole feels scattered, the next step is usually not a new tool — it’s a shared structure and direction.

→ Book a 30-minute conversation — we’ll go through where you stand and what the next step could be.

---

From Astu Labs, Lenni Laukkanen, Jouni Manninen and Janne Pihlajamäki took part in this case. Meet the Astu Labs team here.