Computers and business systems are used more frequently in a world where IT gradually is taking over. This is also true when it comes to the real estate valuation process where actors on the Swedish real estate market are using automatic valuation systems for real estate appraisal. But the real estate agents have not yet seen the charm in using these automatic valuation systems for real estate appraisal. Can the real estate agents also use the automatic valuation systems? We developed our three main subjects: the valuation process, an automatic valuation system and the real estate market with assistance from scientific articles. Through these articles we will present a background to real estate appraisal in general. We choose a qualitative approach where we interviewed eight real estate agents and complemented this by conducting a telephone interview with a business area manager at a private valuation institute. The result of this paper shows that the real estate firms rely on data from earlier purchases and that they use different IT-based tools in order to make an accurate appraisal. An automatic valuation system with extended information can be very helpful for the real estate agents depending on which part of the country the real estate agent works in. However the automatic valuation system can never replace the ocular inspection. In neighbourhoods where the houses look the same there’s a greater need for an extended base of information through an automatic valuation system. In areas with unique locations and houses that are difficult to compare the demands are higher on the content of the system to be correct and updated. In this way the automatic valuation system can be used in the housing market and fulfils a need in both types of areas as a complement to the ocular inspection.

Investment decision in the property market is closely connected to property valuation. Thus, accuracy of valuation results and deep analysis of the market is essential. Artificial Intelligence (AI) models have been successfully adopted in different fields and markets. However, the real estate market is typically lagged in time to adapt to these changes. Swedish commercial property market arrangements are characterized by increasing confidentiality of certain data types. As a consequence, the adoption of the AI valuation models in the Swedish commercial property market is slowed down.  This study aims to bridge the gap in existing research by focusing on the market actor’s behavior in relation to market development and exploiting the capabilities inherent in adopting AI models in commercial property valuations.  The qualitative approach based on interviews with experts has been used to achieve the main objective of this study. Results suggest that the AI valuation models used on commercial properties are applied on valuation data and not on real transaction data. Analysis covers different aspects including data challenges and its disclosure, the role of government authorities, market and data perspectives of AI application on property valuations. A framework on AI implication in property valuation in different time horizons presented in this study will help to overcome data challenges and improve transparency of valuation results. This study is beneficial to various actors in the property market, including government authorities, investors, valuers and researchers. / Investeringsbeslut på fastighetsmarknaden är sammankopplat till fastighetsvärdering. Således är noggrannhet i värderingsresultat och en djup marknadsanalys nödvändiga. Artificiell intelligens (AI) modeller applicerades framgångsrikt inom olika områden och marknader. Fastighetsmarknaden är dock försenad i tid för att anpassa sig till dessa förändringar. Svenskt kommersiellt fastighetsmarknadsarrangemang är känd för ökad sekretess för vissa datatyper. Som en följd av detta minskar adopteringen av AI-värderingsmodeller på den svenska kommersiella fastighetsmarknaden. Denna studie syftar på att fylla i gapet i befintlig forskning genom att fokusera på marknadsaktörens beteende i förhållande till marknadsutveckling och utnyttja de möjligheter som ligger i adopteringen av AI-modeller i kommersiella fastighetsvärderingar.Den kvalitativa metoden baserad på intervjuer med experter har använts för att uppnå huvudmålet för denna studie. Resultaten tyder på att AI-värderingsmodellerna som används på kommersiella fastigheter tillämpas på värderingsdata och inte på transaktionsdata. Analysen täcker olika aspekter, inklusive datautmaningar och dess avslöjande, myndigheternas roll, marknads- och dataperspektiv för AI-tillämpning på fastighetsvärderingar. Ett ramverk för AI-implikationer i fastighetsvärdering inom olika tidshorisonter som presenteras i denna studie kommer att hjälpa till att överkomma datautmaningar och förbättra transparensen i värderingsresultaten. Denna studie är nyttig för olika aktörer på fastighetsmarknaden, inklusive myndigheter, investerare, värderare och forskare.