Uppsats

Från AI till anställning : AI inom rekrytering och balansen mellan effektivitet och rättvisa

Kandidat-uppsats

Umeå universitet/Institutionen för informatik

Publicerad: 2025

Språk: Svenska

Nyckelord

klicka för att söka

Sammanfattning

Artificial intelligence (AI) has a fairly large influence on the world as we know it today, and it just keeps on growing. It is especially big when it comes to recruitment, where a whole 28 precent of Swedish companies are already using it in some capacity. This study research Swedish HR-managers thoughts and experiences with AI in the recruitment process. It especially focuses on how they use the AI and what impact, both good and bad, it can have on recruitment. A big factor that could potentially affect AI is the coming EU-legislation, the AI-act. These coming regulations are the basis for several points in this study. And with this in mind this study focuses on some key element based on the AI-act such as transparency, Bias and Human-AI-relations. The conclusion of this study points to several common advantages of AI tools in recruitment, including time-efficiency, general task relief and being a tool for inspiration. Some of the respondents saw AI as a helpful part in reducing bias, while others instead perceived it as reinforcing existing biases. A view that was shared among respondents was that AI worked well as a complement to the human workforce but could not be fully trusted enough to replace it. Transparency was another aspect considered important, particularly when I came to the later parts of the recruitment process, such as the selection phase, whereas in earlier parts of the recruitment process was seen as less important and dependent on context. Many of the respondents showed little or no awareness of the upcoming EU legislation and considering that AI in recruitment is categorized as a hight-risk in the AI-act, it may be pivotal for these organizations to familiarize themselves with these new legislations.

Information

Lärosäte / institution
Umeå universitet/Institutionen för informatik
Publiceringsdatum
2025
Uppsatstyp
Kandidat-uppsats
Språk
Svenska

Utforska vidare

Liknande uppsatser

Uppsatser med liknande ämnen och nyckelord.