Sep 18, 2026 · 35 min
Artists challenge AI’s right to train on their work
When AI scrapes your art
The dispute over scraped artwork tests whether creators can secure consent, credit, and compensation as generative AI scales.
- 1Artists face lost commissions when AI systems imitate styles built from scraped work.
- 2Data poisoning offers a temporary defense while courts and lawmakers debate fair use and attribution.
- 3Licensing, compensation, and restrictions could reshape AI without banning the technology outright.
Don't miss
Kelly McKernan describes discovering that more than 50 paintings had been scraped into AI training data and seeing commissions disappear.
The brief
Kelly McKernan discovered that more than 50 paintings had been scraped into AI training data, enabling users to generate images resembling their distinctive style.
As clients turned to faster, cheaper AI-generated images, McKernan lost commissions and confronted the difficulty of detecting scraping or securing consent and payment.
Artists can use anti-AI tools and poisoned artwork to disrupt training, but the episode treats those tactics as temporary defenses rather than lasting policy.
NYU law professor Jason Schultz explains why fair use will shape many AI copyright disputes, while distinguishing claims against companies from potentially infringing user behavior.
The discussion weighs licensing, attribution, compensation, and product restrictions, asking whether settlements could constrain AI without banning it outright.
Books & mentions
Some links are affiliate links — PodLume may earn a commission if you buy.