AI, power, ethics and legitimacy – what does Nordic AI leadership require?

In the previous post in my AI series, I wrote about the controls that determine where the power lies when AI influences decisions. The next level is about ethics. When decision-making power is shifted from people to systems, not only a governance issue arises, but also a legitimacy issue.

Who is perceived as responsible when an algorithm prioritizes one customer over another? Who explains why a risk model gives different results? What happens when a dynamic pricing model differentiates prices based on willingness to pay - and the margin increases, but the experience of fairness weakens? What happens to the trust if the decision is effective, but is perceived as unreasonable?

Ethics in AI management is not primarily about regulations, but about how power is exercised and perceived. Here we see clear regional differences.

In the US, speed, scaling and market power are rewarded. Innovation is rolled out quickly, and corrections often take place afterwards. In parts of Asia, high technological progress is combined with strong central management, where direction and responsibility are defined from above.

In the Nordic countries, we stand in a different tradition. High institutional trust (OECD, World Values ​​Survey) and strong employee rights mean that legitimacy cannot be taken for granted. EU regulation, including the AI ​​Act, reinforces the requirements for accountability and explainability. Decision-making systems must withstand legal and social testing.

The debate surrounding the Norwegian Oil Fund's AI investment illustrates this field of tension. Ambitions to boost productivity and save billions are understandable, but raise questions about systemic risk, loss of institutional competence and "alpha decay" when many people use similar models. The question is not whether AI should be used, but whether the pace is accompanied by sufficient risk assessment and ethical grounding.

It can be experienced as a pace handicap, but also as a competitive advantage. Businesses that combine technological progress with ethical grounding will stand stronger over time. Without legitimacy, trust is weakened, and thus room for action and competitiveness.

AI management is therefore not only about governance, but about the responsible exercise of power. It is not just about what the systems can do, but about what they should do and how it is explained.

The real test of maturity is not how advanced the model is, but whether the organization can withstand the consequences of the decisions it influences, even when they are challenged.

The next step in the series is about the manager's judgement. When AI becomes part of the decision-making basis, it is no longer enough to understand the governance structures. Managers must understand how models respond to context, how hallucinations occur, and how recommendations should be critically assessed.

AI judgment is not a niche technical skill. It is a leadership skill.