Artificial intelligence helps attribute Rembrandt with clear hints while humans stay in charge
Identifying a painting’s maker is long and hard; a computer tool shows clues without taking away human judgment.
- Publication
- PNAS
- Stage
- Journal article
- What we read
- Summary of the abstract
- Authors
- Andrés Arias-Rosales, Arthur K. Wheelock, J. Kenneth Moritz, Philip R. LeDuc, Jonathan Cagan
- Universities and research institutions
- Carnegie Mellon University, Universidad EAFIT, University of Maryland, College Park
Institution metadata: OpenAlex record ↗
What they did and found
Attribution is a long-standing puzzle for paintings. The study uses artificial intelligence to show clear, explainable clues while people still judge, noting when the system is confident and when it isn't.
Why it matters
Experts can use the hints to fuel discussion while keeping final judgment with humans and avoiding overreliance on scores.
The topic is hard and cultural. The approach adds readable visual clues that people can understand, keeping a human in the loop when deciding attribution.
The finding shows strong results when the system is confident, while stressing the need for human judgment in every call and for careful calibration to avoid overconfidence.
What remains uncertain
Abstract-only scope; preprint status noted where applicable; not validated by full paper. Reading did not assess full methods or outcomes beyond the abstract.
Original sources · 1
- Rembrandt through the eyes of the machine: Toward explainable Human–AI collaboration in painting attribution ↗PNAS · 2026-10-26
Check the original paper for its authors, methods, version and access terms.
What is your take?
Ask a question, add useful context or share a different perspective. Keep the conversation respectful and grounded.