AI, Art and the Question of Value

Gallery research and collector perspectives raise different questions about usefulness, artistic intent and trust in an increasingly automated creative landscape.

Arts & Business Institute

A tool can become useful before people agree on the value of what it produces. Two recent accounts of AI in the art world offer a starting point for examining that distinction—and for asking better questions about technology inside organisations.

Two different kinds of evidence

Artsy’s February 2026 survey received responses from more than 300 gallery professionals. Its findings describe the use of AI for operational tasks alongside continuing doubts about AI as an artistic medium. The sample includes Artsy partner and non-partner galleries; it should not be treated as a census of the art world.

IEEE Spectrum’s reporting adds selected collector, artist and curator perspectives on what makes AI art meaningful or collectible. It brings questions of process, authorship and transparency into view. These interviews complement the gallery survey, but they do not measure the same population or establish a general level of demand.

The ABI perspective

For leaders considering AI, we would separate three decisions: whether a tool helps complete a task, whether its output meets the required standard, and whether the organisation can explain and take responsibility for using it. A positive answer to one question does not settle the others.

A useful team discussion could compare two outputs without first revealing how they were made. After assessing them, disclose the process and ask whether the evaluation changes. The aim is to identify the criteria people actually use, including accuracy, originality, relevance and confidence in the source.

A question for your next discussion

When production becomes easier, which part of your work deserves more attention: generating an output, judging it, or explaining why someone should trust it?