AI layoffs are considered a realistic possibility by almost half of Danish companies over the next two years, according to a new survey by recruitment consultancy Konsulenthuset Ballisager. But there is still little evidence of a broader decline in employment caused by artificial intelligence in Denmark. In fact, employment has so far increased in most of the occupations considered particularly exposed to the technology.
According to the survey, reported by TV 2, 46% of Danish companies consider it possible that they will dismiss employees within the next two years because some of their tasks are taken over by AI.
Morten Ballisager, owner and CEO of Konsulenthuset Ballisager, described the proportion as worryingly high, while cautioning against treating it as evidence that mass unemployment is approaching.
The distinction is important. The survey measures what companies believe could happen over the next two years. The available data measures what has happened so far. And for the moment, the two tell rather different stories.
AI layoffs have not yet appeared across the Danish labour market
An August analysis by the Economic Council of the Labour Movement (Arbejderbevægelsens Erhvervsråd, AE) examined employment in nine occupational groups considered among those most exposed to artificial intelligence.
Employment increased in seven of the nine groups. Taken together, the nine recorded growth of around 5%, equivalent to more than 15,000 additional full-time workers.
The result does not demonstrate that AI will have no negative effects on employment. Nor does it show what would have happened to these occupations without the technology. It does, however, challenge the assumption that the jobs most exposed to AI are already disappearing.
So far, technological exposure and employment decline have not been the same thing.
This is also the argument made by Dansk Industri, one of Denmark’s main business organisations. Industry director Andreas Holbak Espersen told TV 2 that some functions will disappear or change, but that the organisation expects the overall effect of AI on employment to be positive.
He also said that, when Dansk Industri asks its members more broadly, companies are not generally reporting that they have begun eliminating positions because of AI.
Danish companies are moving beyond experimentation with AI
One reason expectations may nevertheless be changing is that artificial intelligence is becoming a more ordinary part of Danish workplaces.
A survey by Dansk Erhverv published in February found that 70% of its member companies use AI tools, compared with 44% in 2023.
Among companies with more than 250 employees, the proportion reached 96%. Of the businesses already using AI, 73% said the technology had made their workflows more efficient.
This does not necessarily mean fewer employees. Greater productivity can reduce the amount of labour required for an individual task while simultaneously allowing companies to produce more, change workers’ responsibilities or create other functions.
But there are now concrete examples of companies expecting AI to affect the size or composition of their workforce.
Nordea and Netcompany show how the change could happen
In March, Nordea announced restructuring measures expected to affect approximately 1,500 employees across the group in 2026 and 2027.
The bank did not attribute those changes to AI alone. In its official announcement, it said its future workforce would be shaped by a combination of greater Nordic integration, AI and process optimisation.
Nordea expects to employ fewer people in the future, while simultaneously changing the skills it needs. The bank said affected employees could be offered reskilling, upskilling and internal opportunities.
A similar distinction is important in the case of Danish IT company Netcompany.
The company has set aside DKK 148.8 million (€19.9 million) for dismissals and organisational adjustments associated with its changing workforce and the introduction of agentic AI across the group, according to Finansforbundet.
Netcompany has not disclosed how many employees could be affected.
CEO André Rogaczewski has also rejected the description of the changes as a conventional round of layoffs. The company still expects its total workforce to grow, but more slowly than its revenue, while some traditional consultancy and development roles may become less necessary and demand increases for other skills.
These examples illustrate why measuring the employment effects of AI is difficult. A company can reduce some roles, hire for others and increase output at the same time.
Junior workers could face a particular problem
Another indication of how AI could reshape employment comes from PwC’s CEO Survey 2026.
Among the 77 Danish CEOs who participated, 43% said they expected the number of junior-level employees to decrease because of AI. The corresponding figure was 23% for middle-level employees and 5% for senior positions.
The survey mainly reflects the expectations of executives at some of Denmark’s larger companies and should therefore not be read as representative of every Danish employer.
It nevertheless points to a potentially important consequence of AI adoption even if mass unemployment does not materialise.
Many entry-level positions contain the routine tasks that allow younger employees to acquire experience before taking on more complex responsibilities. If companies automate a significant share of those tasks, they may need fewer junior employees without necessarily reducing their overall workforce by the same proportion.
The result could be a change in the structure of employment rather than simply fewer jobs.
Denmark is starting from a strong labour market
The transition is also taking place while the Danish economy and labour market remain relatively strong.
A new AE economic forecast expects GDP to grow by 4.2% in 2026, after growth of 3.5% last year. Employment has increased by around 300,000 people since 2021, according to the organisation.
AE expects unemployment to increase by around 6,000 people in 2026, but from a historically low level.
That gives Denmark more room to absorb changes in the composition of employment than it would have during a period of widespread job losses.
It also helps explain why the two pictures of AI and employment can currently coexist. Companies can expect to need fewer people for certain tasks while the economy continues creating jobs elsewhere.
The Ballisager survey therefore says something significant about how quickly employer expectations are changing, but it does not yet amount to evidence of mass AI-driven job losses in Denmark.
For now, the more visible effect is a transformation in what companies expect employees to do. Whether that eventually produces fewer jobs overall remains an open question.
