| dc.description.abstract | This thesis investigates firms’ adoption of artificial intelligence technology and its effects on
firm performance. Due to the rapid technological advances of artificial intelligence, this field
of research is still largely unexplored, especially in Norway. The research question guiding this
study is: "Which Norwegian firms are adopting AI, and how does this adoption affect firm
performance?"
To estimate and analyze the adoption of AI, we employ web scraping-based methodology and
conduct textual analysis of company websites. This approach taps into an underutilized source
of information for research purposes. However, it is important to note potential concerns related
to endogeneity due to the limitations of our cross-sectional data.
Based on our analysis of approximately 53,000 Norwegian companies, we find that several
factors influence the likelihood of having adopted AI. Generally, companies located in urban
areas, startups, those with more employees, and those with male CEOs are more likely to have
a positive AI Know-how score. However, we also identify nuanced variations. Specifically, the
number of employees positively affects the AI score in the computer programming industry but
does not exhibit the same relationship in the advertising or transportation sectors. This suggests
that the number of employees positively relates to AI adoption in industries where value is
driven by employee capabilities and complementary resources.
Furthermore, within our sample, we observe that 2.7% of Norwegian firms utilize AI
technology. Notably, the "Telecom, IT, and Media" industry group exhibits the highest
proportion of positive AI Know-how scores, with 11.87%. We find that firms adopting AI
experience a lower return on their assets, a lower operating margin, and lower sales per
employee than their non-adopting counterparts, indicating that today’s Norwegian firms' AI
capabilities do not lead to higher performance. However, we see increased sales growth in AI
adopters, indicating a focus on future growth for these firms.
These findings contribute to the growing literature on AI adoption and offer insights into the
Norwegian context. Additionally, our thesis can serve as a valuable starting point for future
research employing similar methodologies. | en_US |
