LO
CapabilityCognitive automation2026-01-17
A machine-learning model matched the accuracy of 13 real estate experts in much less time, and experts working with it were most accurate, a Vienna study found
Property appraiser / valueroccupation page →
Event date / reported
2026-01-17
Evidence stage
CapabilityA demo, benchmark or paper shows the task can be done. Updates what the technology can do — not what employers will do.
Tasks this bears on
Analysing sales and estimating value
Choosing comparable sales, adjusting for differences and reaching a value for a lender or client.
Automating✓ Evidence-backed
Where this applies
Vienna, Austria. A model trained on 21,736 transactions was compared with 13 experts valuing newly built apartments sold in 2023, under limited information, the experts' usual methods, and collaboration with the model. The model achieved accuracy comparable to the experts while greatly reducing the time needed, and experts working with the model achieved the highest accuracy. A small experiment in one city; it does not measure use in practice.
What this means
A model can value standard flats about as well as experts, faster, and experts do best with it.
What it does not yet show
One experiment with 13 experts in one city.
What you can check
Open PubMed record 41547978 and find "accuracy comparable to that of experts".
Does it change the assessment?
No — and this stage does not move it either. A "Capability" record is real evidence, but it does not upgrade a task judgement on its own. The 1 linked judgement above stand where they were.
Source
Kmen C et al. — Evaluating human-machine collaboration through a comparative analysis of experts, machine learning, and hybrid approaches in real estate valuation, Scientific Reports (published 17 January 2026; PubMed 41547978) · verified 2026-10-01 · Claude (VOLO agent) · interpreted 2026-10-01 · Claude (VOLO agent)
Primary source — published by the party that did this, or the authority of record. No co-signature needed.