Recruitment tools are great. They take a lot of work off the hands of HR departments. In some cases, matching tools even suggest suitable people. Computers are also emotionless, which means they are neutral – so to say. But herein lies the crux of the story. Recruitment tools are programmed by people, and existing personnel files serve as the data basis. So when the programmers feed in personnel files from previous employees, similar profiles are suggested again. But is this what we want in the age of diversity? Wouldn’t it make more sense if the computer suggested different profiles, e.g. people of foreign origin, young people, older people, women and men? Yes, but unfortunately this does not happen. The machine learns from the past and, depending on the programming, the algorithm “aggravates” in the direction of reinforced stereotypes. In the medium term, this would then lead to an even more pronounced occupational segregation of the labour market and greater inequality of opportunity.

Let’s take an example: A woman with a STEM university degree becomes a mother at 33 and takes a two-year sabbatical. After that, she wants to return to work. STEM degrees are in demand, she thinks to herself, so she sends out job applications, but immediately receives a rejection from all the larger companies. A look behind the scenes shows: The two-year break is recognized by the system, evaluated negatively, and the application is thus immediately filtered out.

It takes machines and humans

What does this mean for companies? It is crucial that HR managers look at the individual steps in the recruitment process and consider, from a diversity perspective, where the use of artificial intelligence makes sense and where the experience of recruitment experts is superior. If you want a more diverse workforce, you can adapt the tools to accommodate more diverse criteria. For example, you can expand the “education” criterion (e.g. from mechanical engineering to include other technical training) or adjust the required years of experience. By experimenting and observing how such adjustments affect the kinds of applications your company receives, you’ll be able to improve the quality of the recruitment process. It’s important to ensure that HR and line managers still make the final decision, supporting the process with artificial intelligence to increase efficiency, but also providing human support.

How to bypass the AI tool

So what does this mean for a person who wants to apply for a job, but has a non-linear career on their CV (e.g., they’ve taken a break, changed careers, or are even over 50)? AI tools are not the only way people get hired. 75% of all job vacancies are filled through personal networks. Therefore: activate your network, update your LinkedIn profile, and let everybody know that you are looking for a new job. Another option is networking events, such as the one we offer at the HSG Career Relaunch conference. Events like these are great opportunities for job seekers to meet with HR managers and also apply for programs like the UBS Career Comeback Program. And last but not least, continuing education courses also help to expand one’s network. Three-quarters of the women who completed the “Women Back to Business” program have either found a new job after their career break, successfully repositioned themselves, or taken on a position with higher responsibility.

About the author(s)


Executive School of Management, Technology and Law


Get the latest articles directly to your inbox.

Newsletter [EN]

Share article

More articles

  • International Award for Research on Digital Transformation for EMBA Lecturer Prof Dr Jan Marco Leimeister

  • University of St.Gallen among the top 20 worldwide in the Financial Times Executive Education Ranking 2024!

  • Research project by Prof Dr Damian Borth in the field of machine learning is supported by the SNF

  • ‘Elite Quality Index 2024’ with Prof Dr Tomas Casas places Switzerland in second place

  • Karolin Frankenberger published a new book!