Speaker: Stan Berger, independent researcher, historian, art critic.

The webinar opened with Stan Berger dismantling the common fear that AI will replace human creators. Instead, he framed generative models as “collaborative brushes” that expand the artist’s vocabulary, not diminish it.

Berger devoted a substantial part of the session to the unresolved question of authorship: while current copyright laws do not recognise an AI as an author, the prompts and curation behind the output may qualify for neighbouring rights protection. He then walked the audience through several landmark lawsuits filed by visual artists against AI developers, proposing a practical “opt‑in” licensing model that would let creators voluntarily offer their works for training while retaining control over commercial use.

The most provocative moment came when Berger presented a real‑world conservation case – a neural network had been used to reconstruct missing fragments of 15th‑century frescoes, raising fierce debate over whether such algorithmic intervention violates the integrity of the original masterpiece. He concluded that AI is already embedded in our cultural ecosystem, and the real challenge is to build transparent governance frameworks that safeguard human intuition while harnessing the technology’s analytic power.