The 4% Problem – Why Nordic AI Investments Are Not Yet Delivering Business Value

Signal 1: Investments are outpacing strategic clarity

The most revealing figure in this quarter’s data is not the level of investment. It is the leadership figure.

Only one-quarter of Nordic CEOs are actively involved in their company’s AI strategy. Globally, that figure is closer to 49 percent. The Nordics are therefore operating at roughly half the level of the global average. The consequences are visible directly in the value gap.

When AI strategy is primarily owned by the technology function, it risks becoming a technology issue. The focus then shifts towards tools, platforms and vendors. The questions that actually determine whether AI creates business value – which problem should be solved, what success looks like, and which initiatives contribute to business objectives – are rarely asked.

Deloitte’s Nordic data makes this clear: 40–50 percent of Nordic AI investments go towards off-the-shelf tools added on top of existing processes. The corresponding figure globally is 8–11 percent.

This is not a budget issue. It is a result of getting the order wrong. Technology should not drive the strategy. The strategy should guide the technology, with technology acting as an enabler. When technology takes the lead, organisations tend to invest in tools before they have defined what they want to achieve with them. Productivity may improve, but the intended transformation does not happen.

The signal:

Strategic clarity is becoming the biggest gap in Nordic companies’ AI efforts. The organisations that will close the value gap first are not those with the most advanced AI solutions. They are the ones whose leaders have the clearest answer to the question: Why are we doing this?

Signal 2: The skills gap is an organisational challenge, not a recruitment challenge

A year ago, 61 percent of Nordic companies said they felt strategically prepared for AI. Today, that figure has dropped to 43 percent. Strategic readiness has declined by 18 percentage points in twelve months, even as AI investments and ambitions have accelerated. The more companies have invested, the less prepared they feel.

The skills figure is even more concerning. Only 14 percent of Nordic organisations believe they have the right capabilities in place.

The instinctive response to a skills gap is to recruit. But the shortage is not primarily about technical expertise. The hardest roles to fill are not AI users or even people who develop AI solutions. They are the people who can make AI work across the organisation: those who can coordinate AI systems across workflows, govern them responsibly, and continuously improve solutions as needs evolve.

This is exactly where many skills strategies fall short. An organisation can hire the right person and still prevent them from making an impact. If decision-making authority sits several levels away from the work, if approval cycles take weeks instead of days, and if teams are designed for stability rather than continuous improvement, the organisation cannot benefit from the capabilities it has brought in. When an organisation is built for a different era, it limits the very capabilities it has just recruited.

Many Nordic organisations have made precisely this mistake. They have introduced AI into structures designed for a slower pace and then wondered why newly hired experts are not delivering the expected results. The tools are sophisticated. The structures resist change. Skilled employees are forced to compensate for organisational limitations.

The signal:

Skills readiness and an organisation’s ability to change are not two separate challenges. They are the same challenge viewed from different perspectives. Organisations that treat them separately will continue hiring people into structures that limit their ability to create value. Those that close the gap will develop both the capabilities and the ways of working required to make AI deliver.

Signal 3: Trust becomes a competitive advantage

In a 2026 Usercentrics study, 47 percent of consumers said that, during the previous six months, they had made at least one change affecting their purchasing decisions due to concerns about how their data is used in AI services. This means that almost half of consumers have either reduced their purchases, switched providers, or ended a relationship because of AI-related trust concerns.

This issue is particularly important in a Nordic context. Nordic societies are characterised by high expectations around transparency, long-term responsibility, and institutions acting in a trustworthy manner. When AI challenges these expectations, the consequences can be significant.

At the same time, many organisations still view trust primarily as a compliance issue rather than a strategic brand consideration. In August 2026, the transparency requirements in Article 50 of the EU AI Act will come into effect, requiring organisations to inform users when they are interacting with AI systems and to label AI-generated content. Only 3 percent of Nordic companies say they feel fully prepared for these requirements.

This creates a choice. Organisations that see compliance as the finish line will meet the minimum requirements and move on. Those that treat transparency as a brand principle will build something more enduring: the kind of trust that survives mistakes, encourages customers to recommend a brand, and creates loyalty that competitors cannot easily replicate.

In a world where AI-generated content is everywhere, the most valuable thing a brand can offer is genuine human trust. It is not just a perception. It has a direct impact on business outcomes. Trust is becoming a critical competitive factor.

The signal:

Brand trust and AI governance are becoming increasingly connected. Organisations have an opportunity to make transparency part of the brand experience and build stronger customer loyalty.

Signal 4: Agentic AI is here before organisations are ready

Agentic AI – systems that do not just respond to instructions but can also plan, execute, delegate and adapt across complex workflows – is rapidly moving from experimentation to practical use. Globally, 79 percent of companies report that AI agents are already being used or tested within their organisations. The agentic AI market has surpassed nine billion dollars, and 88 percent of business leaders worldwide plan to increase their AI budgets due to agentic AI initiatives.

In the Nordics, AI adoption is progressing more cautiously. 54 percent of Nordic leaders say they are experimenting with AI agents, while one-quarter are still observing and planning. Nearly 60 percent of Nordic companies allocate less than 5 percent of their AI budgets to agentic AI initiatives. BCG describes this as a risk of a local AI value bubble: if the gap between Nordic investment and Nordic returns remains as agentic AI scales globally, the region risks falling behind – not because it lacks ambition, but because organisations wait too long to adapt.

The challenges around readiness are structural. Agentic AI requires something most organisations have not yet built: a clear knowledge architecture. When an AI agent executes multi-step workflows, creates content, briefs teams and provides recommendations that influence business decisions, the quality and reliability of its output depend on the knowledge, principles and context available to it. Without a defined strategic framework, clear brand language and documented decision-making principles, agentic AI risks producing fast, generic results at scale. It amplifies what already exists. If the foundation is underdeveloped, amplification makes things worse, not better.

The organisations that will gain the most value from agentic AI are those that have done the foundational work: established strategies, defined language, documented principles and built knowledge systems that AI agents can use as the basis for their decisions.

The signal:

Readiness for agentic AI starts with knowledge architecture and only then moves to technology. Organisations building their knowledge systems today are not simply creating documentation. They are building the foundation required for the next generation of ways of working.

Signal 5: Activity metrics are hiding the value gap

Nordic organisations are measuring the wrong things. The consequences are starting to show.

79 percent of Nordic companies report improved efficiency from AI. That looks like success until compared with revenue figures: only 18 percent see increased revenue as a result of their AI investments, despite 75 percent expecting it. The gap between reported efficiency gains and actual business value is a clear signal.

One important explanation is how organisations measure their results. In most businesses, the easiest things to count are the ones being measured: time saved, automated processes and implemented tools. These are activity metrics. They show what the organisation has done, but reveal very little about whether the business has actually improved.

This also influences what organisations prioritise. A business that measures time saved will optimise for time saved. An organisation that measures business outcomes will make different priorities, invest differently and set much higher standards before approving new initiatives. What you measure determines what you prioritise.

BCG’s analysis points to the same structural challenge. Organisations are allocating most of their resources to initiatives that generate marginal productivity improvements while underinvesting in initiatives that transform workflows and create new business capabilities. When measurement rewards the former, the latter rarely gets prioritised.

The companies that will achieve the greatest digital success in the coming years will not be those with the most AI tools. They will be those that measure business outcomes such as revenue, retention, customer lifetime value and time to market. Results that a board can understand, customers can experience and competitors cannot easily replicate.

The signal:

AI does not create business value simply because it is being used. The organisations that succeed are those that measure outcomes in terms of revenue, customer value and competitiveness – not just hours saved and tools implemented.

What this means overall

Together, the five signals describe the same underlying challenge from different perspectives. Strategy is not guiding technology. Organisations are not built for the pace that AI requires. Trust is not keeping up with the development. The knowledge foundations are not ready for the next generation of AI. And companies are still measuring activity rather than business outcomes.

Each of these challenges is significant on its own. Together, they show that the Nordics have been quick to adopt AI, but considerably slower in building the organisational capabilities required for AI to create long-term business value.

The solution is not more technology. It is about building a clear strategy, flexible ways of working, strong trust, a well-designed knowledge architecture, and measurement frameworks that connect investments to business outcomes.

The data from Q2 2026 shows that the cost of starting in the wrong place is increasing. AI tools are becoming cheaper and more accessible at a rapid pace. Competitive advantages will therefore not come from the technology itself, but from an organisation’s ability to use it effectively.

This is where Nordic companies have an important advantage. Nordic business culture is characterised by trust, transparency and long-term thinking – qualities that are becoming increasingly important as AI becomes an integral part of business operations. The question is whether companies can build on these strengths, or whether technological development will move faster than their organisations can adapt.

This text has been translated with the help of AI. We apologize for any ambiguities or inaccuracies that may have occurred during the translation process.