What happened
AI algorithms have been introduced to assist with art attribution, analysing brushstrokes, colour palettes, and stylistic elements to identify artists and authenticate artworks. This technology processes vast datasets, including authenticated works and known forgeries, to recognise formal patterns. While AI supplements human expertise by identifying patterns potentially missed by experts, it exhibits inaccuracies, with dating errors sometimes exceeding a century, and does not account for non-visual factors like historical context or provenance.
Why it matters
The integration of AI into art attribution introduces an operational constraint by creating a visibility gap in non-visual authentication factors, such as historical context and provenance. This increases exposure to potential misattributions and significant dating inaccuracies, given AI's reported errors exceeding a century. Authentication teams and collection managers face a higher due diligence requirement to ensure comprehensive analysis, as AI's purely visual assessment necessitates robust human oversight for critical non-visual data.
