Study: AI hits young job seekers hardest

Years of discussion about AI’s impact on the labour market have led to contrasting visions of the future — some optimistic, others deeply dystopian.

There has been little solid data to back up these scenarios however. Until now. A study, ‘Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence’, authored by three researchers at Stanford University in Palo Alto, California shows how AI impacts job opportunities by using data on millions of employees from ADP, the largest United States payroll provider. This has given the researchers a large and reliable dataset.

The conclusion is that one particular group of highly educated workers is especially vulnerable. In the US, there has been a 13 per cent drop in the hiring of young employees aged 22–25 within customer service and IT, and this trend is linked to the introduction of artificial intelligence.

Danish white-collar workers

The question is whether the same group of highly educated workers in Denmark should be worrying about their future job prospects.

Artificial intelligence has the potential to accelerate the transformation of the entire labour market exponentially

Christian Lyhne Ibsen, associate professor at the Department of Sociology

The University Post asked associate professor Anna Ilsøe from the Department of Sociology at the University of Copenhagen (UCPH) about the prospects for new graduates finding and keeping a job in the Danish labour market when they are up against AI:

»The study shows that the hiring slowdown affects office-based jobs, which also include university graduates. But anyone who works with computers could be impacted.«

Anna Ilsøe has researched digital labour markets at the Employment Relations Research Centre (FAOS) for many years, and she and her labour market researcher colleagues have already observed AI-related changes in parts of the Danish labour market.

»It appears that the jobs most affected by AI here are those in programming and IT, customer service, office-based jobs, and finance,« she says.

AI will affect Europe differently

Anna Ilsøe reckons that Europe will tackle the AI challenge in a different way than the US in the coming years however.

In the US, Amazon recently laid off thousands of office workers, but no drivers or warehouse staff were let go — as their manual, practical work could not be replaced by AI. The remaining office staff, on the other hand, were expected to deliver more per paid working hour, as AI made it possible to automate many of their tasks, thereby increasing productivity requirements.

But according to Anna Ilsøe, we are unlikely to see the same thing happening in Danish companies — at least not in the near future.

»In a European and Nordic context, companies will stop hiring new employees instead, with existing staff being retrained and upskilled so they can work alongside AI,« she says, and continues:

»Experienced employees are also unlikely to be at high risk of dismissal, because they know how to solve the problems that AI cannot handle, and they know how AI can help the company tackle challenges.«

Another question is how AI will impact the labour market in the long term — in Denmark, the Nordic countries, and Europe.

»We don’t know yet, but you can easily imagine that in the future we might move towards something resembling what we’re seeing in the US today,« she says.

Knowledge-intensive jobs lost in new ways

Anna Ilsøe’s research colleague, associate professor and centre director Christian Lyhne Ibsen at FAOS at the Department of Sociology, has also read the Stanford University paper, which he finds thought-provoking.

»It’s not surprising that new technology leads to job losses in Denmark. What is surprising is that employees in knowledge-intensive positions may soon start losing their jobs in a different way than before,« he says, and adds:

»The types of knowledge-based tasks that are also standardisable — which we have so far outsourced to Eastern Europe or India — will no longer be sent away. Instead, they will be handled by artificial intelligence, and this will have consequences for how companies structure themselves.«

Shields senior roles, sidelines entry-level staff

In the past, it was mainly manual and unskilled industrial jobs that were offshored to low-wage countries. This resulted in job losses in high-wage countries like Denmark.

But according to Christian Lyhne Ibsen, the nature of offshoring has changed. Now it is also jobs that require high knowledge levels that are being offshored from high-income to low-income countries.

He uses legal work as an example. Tasks in this area are being relocated from the US to countries like India, where language skills and legal expertise are competitive.

This has contributed to the job structure in US law firms switching from a pyramid shape — with many low-paid employees at the bottom and a few high-paid staff at the top — to a diamond shape: narrow at the top and at the bottom, but wide in the middle. The shift takes place mainly at the bottom, and this is not good news for newly graduated law students looking for their first job.

»With AI, companies can automate knowledge work and, in the process, cut labour costs. There are still only a few people at the top, but far fewer new employees are being brought in at the bottom levels. Experienced employees in the middle can continue to move up the ranks however,« says Christian Lyhne Ibsen.

Replacement jobs for AI

The new diamond-shaped structure can also be explained by the fact that law firms still need experienced mid-level employees to verify and sign off on all the legal specifications produced by AI.

Beyond the individual examples, AI could also pose a challenge simply due to the speed in which the changes are happening.

»Automation and offshoring have for a long time led to job turnover when jobs are relocated abroad, are taken over by machines, or are replaced by new jobs. But artificial intelligence has the potential to accelerate the transformation of the entire labour market exponentially,« says Christian Lyhne Ibsen.

He has been thinking about how AI might shape the future:

»The big question is whether we can come up with new job roles that only humans can do — fast enough — so that artificial intelligence doesn’t end up creating mass unemployment«.

This article was first written in Danish and published on 13 January. It has been translated into English and post-edited by Mike Young.