Uncertainty about upcoming EU AI rules is slowing AI development in Nordic companies as the new regulatory framework nears full implementation.
All parts of the EU AI Regulation will soon be fully applicable, yet a new survey of Nordic companies shows significant work remains. Legal and regulatory uncertainty is causing many AI projects to be delayed or halted altogether.
An international compliance survey from NAVEX reveals persistent challenges for Swedish and Nordic businesses as they prepare for the AI legislation. Since the regulation came into force, the EU’s AI framework has been applied gradually; the process is expected to be complete by early August 2026.
Only 30 percent of Nordic respondents say they are fully prepared to comply with the upcoming AI rules. Meanwhile, 60 percent report they are actively working on preparations but have not yet reached compliance.
Just 26 percent consider their company’s internal policies for safe and responsible AI use to be very clear. Nearly as many — 24 percent — find those guidelines unclear, generating uncertainty about how AI should be used in daily operations.
According to Vincent Vacher, Europe North Regional Leader at NAVEX, there is clear room for improvement across Nordic business. He emphasizes that creating clear local frameworks and dedicating resources to staff training are vital next steps. Without those measures, it becomes harder to realize real commercial value from AI initiatives, which can produce bottlenecks and costly dead ends.
Over the past 12 months, 55 percent of respondents reported that one or more AI projects were stopped. For 20 percent of companies, uncertainty about EU AI rules was a contributing factor to the termination of AI initiatives. Other reasons cited include compliance challenges and ethical concerns, at 18 and 11 percent respectively.
The survey also shows that AI is still used despite unclear internal guidelines. Only eight percent of respondents say they never rely on AI-generated recommendations when they are unsure whether the use complies with internal policies. At the same time, 24 percent say they rely on AI regularly in such situations, and 14 percent report doing so all the time.
The study was conducted by OnePoll in January 2026 on behalf of NAVEX. The Nordic sample includes 100 decision-makers from Sweden, Norway and Denmark, all from organizations with more than 500 employees. The wider survey also covered the UK, France, Germany, Switzerland and the US.
For many organizations the challenge is not technical but governance-related: responsibility, legal clarity and oversight. As EU AI rules become part of everyday business, the need for clear internal policies, training and governance structures grows. Companies that establish these elements early reduce the risk of friction and can convert AI investments into tangible business benefits more quickly.