Nearly half of Finnish job seekers now use AI, survey finds, while employers split on whether it is fair
Use of AI in job hunting in Finland has roughly quadrupled since 2023, according to job search service Duunitori. Employers are split three ways on whether it is fair, and only 26% use AI in recruitment themselves.
Martti Asikainen | 7.10.2026 | Photo by Adobe Stock Photos
Some 45% of active job seekers in Finland have used AI in their job search, up from 37% a year earlier and 11% in 2023, according to a survey by Duunitori and market research firm Taloustutkimus published on 7 October. Duunitori describes itself as Finland’s largest job search service.
That means use has roughly quadrupled in three years. Antti Hyrkäs, employer brand consultant at Duunitori, who is responsible for the company’s recruitment research, said AI is no longer a phenomenon of the future. He described it as a shift that employers need to respond to now.
The survey, titled “Job Seeking in Finland”, is part of Duunitori’s National Recruitment Study. It covered 3,007 Finns and ran from 3 July to 12 August 2026. Duunitori said the results were weighted by age, gender and region.
From CVs to salary talks
Writing remains the most common use. Among active job seekers, 73% use AI to put their skills into words in a CV or application, and 64% use it to tailor documents to a specific job. Another 43% use it to proofread and improve the language. Duunitori said that working on CVs and applications with AI is still the most common use, even as other uses grow.
Use is broadening beyond documents. Some 35% use AI to prepare for interviews and 26% to understand a role or industry. These uses, along with salary expectations and translation, grew fastest relative to last year, according to Duunitori. The release does not give last year’s figures for these uses, so the size of the increase cannot be judged.
“This can improve the applicant’s negotiating position,” said Hyrkäs. According to him, AI now helps job seekers interpret a job’s pay level and their own suitability, not just write applications.
Three camps among employers
Use is uneven. Among knowledge workers, 60% have used AI in their job search, compared with 38% of those in manual work. The share is 62% for university-educated applicants and 34% for those with vocational qualifications.
Women (49%) use AI more than men (41%), especially to describe their own skills. By age, use peaks at 57% among 35 to 49-year-olds. It is lowest among those aged 15 to 24 (40%) and over 60 (32%).
Employers’ views fall into three camps of equal size. Some 28% see applicants’ AI use as permitted and a smart way to gain an edge. Another 28% see it as slightly unfair “cheating”, and 28% as necessary to stay competitive. The remaining 16% could not say.
Job seekers themselves are harsher judges. In Duunitori’s survey last year, 36% of respondents called AI use in job hunting slightly unfair. “Job seekers are their own harshest critics,” Hyrkäs said.
Nearly half of employer representatives (47%) said applications made with AI put pressure on them to reform recruitment. Hyrkäs said the large volume and more even quality of applications make it harder to find the right candidates. Yet Duunitori’s spring survey found that only 26% of employers use AI in recruitment.
AI as sparring partner
Hyrkäs said job seekers can still succeed without AI, but it is becoming a basic tool, much like a search engine. He said he would use it “as a sparring partner, not a substitute”. That means asking it what experience to highlight and what an interviewer might ask, but never letting it invent experience or motivation.
He also warned that a polished application is not the same as a strong one. Real skills, work experience and personality matter more, he said, and AI cannot replace them.
The comparison between job seekers and employers draws on two different surveys, from different years. The employer survey, run from 12 January to 13 February 2026, had 360 respondents, far fewer than the job seeker survey. Differences between the groups may partly reflect timing and sample. The release also does not define what counts as an “active” job seeker.
