Norway wealth fund warns AI as a cost-cutting story can backfire

Nicolai Tangen has warned companies against using AI primarily for job cuts.

Reuters interviewed the CEO of Norway's sovereign wealth fund on April 28. Tangen said companies should use artificial intelligence to "lift everybody up", not just to remove costs. His argument is practical: employees who see AI mainly as a redundancy machine have weak incentives to adopt it.

These are the facts in the Reuters story: the fund invests Norway's oil and gas revenues, owns on average 1.5 percent of all listed stocks globally and holds positions in around 7,200 companies. Tangen said the fund itself is not planning AI-driven layoffs. About half of its 700 employees code their own AI tools.

Why this matters

For Norwegian executives, this is not only an HR point. It is a warning from one of the world's most important owners: AI programmes that start with cost cutting can create weaker internal adoption, stronger employee resistance and lower social acceptance.

The leadership implication is that CIOs and executive teams should measure AI by productivity, quality, response time and new revenue capacity - not only by reduced headcount. Cost benefits will matter, but if they become the main story too early, the organisation may lose trust exactly when it needs learning speed.

Tangen also tied AI to Europe's competitiveness. He said Europe has education, digitalisation and data, but lacks clear technology direction. Reuters also notes that Europe's share of the fund's investments has fallen to 24.8 percent, down from 39 percent a decade ago. That should be read as a capital-market signal, not a model announcement.

Executive consequence

Boards should ask for three things before approving the next major AI programme. First, a benefit model that shows both productivity and capability building, not only FTE reductions. Second, a change plan that makes employees users and co-developers, not just targets for automation. Third, a supplier-risk assessment. Reuters reports that the fund uses one large cloud provider, AWS, and one global custodian, Citibank, for efficiency, while management would like more alternatives for resilience.

That last point is directly relevant to the CIO. AI adoption is not only about models. It also deepens dependency on cloud, data flows, identity, security and operational service providers.

The concrete advice is to run AI as a productivity programme with visible governance. Separate facts from optimism, measure benefits process by process, and include supplier concentration in the risk picture from day one.