How Artists Are Responding to AI
Reading time: 7 minutes
In February 2025, Christie’s opened an auction called Augmented Intelligence - the first major sale dedicated entirely to AI art. Before the first bid was placed, approximately 6,500 artists had signed an open letter calling for its cancellation. They called what was happening “mass theft.” The auction went ahead anyway. And the works sold.
This is where the art world stands in 2026. Not at a tidy fork in the road where artists either embrace AI or reject it, but at a genuinely complicated intersection where the same technology is simultaneously a $900,000 auction result, a class-action lawsuit, a studio tool used by painters in London and Berlin, and an existential question about what creativity actually means.
If you paint for a living - or are trying to - none of this is abstract. It is changing the market you sell into, the tools available to you, and the value placed on the fact that a human being made something by hand.
The artists suing
The legal battle between artists and AI companies is the most consequential conflict in the creative industries since the music industry confronted Napster. Except this time, the defendants are not teenagers with broadband connections. They are some of the most capitalised technology companies in the world.
In January 2023, artist Sarah Andersen and a group of fellow illustrators filed a class-action lawsuit against Stability AI, Midjourney, and DeviantArt, alleging that the companies had used billions of copyrighted images scraped from the internet to train their AI image-generation models - without consent, without credit, and without compensation. The case, Andersen v. Stability AI, is now set for trial in September 2026 in the Northern District of California. If the plaintiffs ultimately succeed in proving their claims, this case could set a transformative legal precedent that will reshape the way AI art is regulated in the United States.
The Andersen case is one of nearly 70 infringement lawsuits filed against AI companies since 2023. The rapid development of generative AI models has given rise to now over 70 infringement lawsuits by copyright owners against AI companies, with cases in different stages of litigation.
The most financially significant outcome so far occurred in September 2025: the Bartz v. Anthropic settlement, in which Anthropic faced a potentially massive statutory damages penalty for downloading millions of pirated copies of works used for training, and reached a reported $1.5 billion settlement. It is the largest AI copyright settlement on record and sent an unambiguous signal to the entire industry about the legal exposure that comes with training models on unlicensed creative work.
For visual artists specifically, the Christie’s auction crystallised the tension. The open letter expressed concern over the potential use of copyrighted material to train generative models, claiming that the auction encourages “AI companies’ mass theft of human artists’ work.” Refik Anadol, whose own work was in the auction, dismissed the criticism as “lazy critic practices and doomsday hysteria.” The divide could not have been more cleanly illustrated - by two artists, both in the same sale, with opposite views on what was happening.
The artists selling
Whatever the legal and ethical disputes, the market for AI art has generated numbers that are impossible to ignore.
Refik Anadol - a Turkish-American media artist who uses machine-learning algorithms to transform vast datasets into large-scale immersive installations - has become the defining figure of the AI art market. His work, which processed millions of data points from the Red Planet, sold for $900,000 at Sotheby’s Riyadh in early 2025, falling comfortably within its $800,000–$1,200,000 estimate. His average lot value reached $227,400 in 2025, driven by the MoMA acquisition of Unsupervised - an installation that processes the museum’s 200-year archive of artworks - which reportedly kept visitors watching for an average of 38 minutes: 76 times longer than the typical gallery glance.
Holly Herndon and Mat Dryhurst, whose AI-generated Embedding Study 1 & 2 - large-format thermal prints built from machine learning data - sold at Christie's for $94,500 against a $70,000–90,000 estimate, represent a different point on the same spectrum: artists who use AI as the primary medium rather than as a tool layered beneath traditional craft.
By 2025, 35% of fine art auctions now feature works incorporating an AI component, according to market data. The AI art market is projected to reach $40 billion by 2033, according to Grand View Research. In spring 2026, DATALAND opened in Los Angeles - the first museum dedicated entirely to AI art, co-founded by Anadol himself.
Sougwen Chung, named one of TIME100’s Most Influential People in AI in 2024, works differently. She paints in tandem with robotic arms she has trained on decades of her own drawing history. The result is neither purely human nor purely machine - it is, as she describes it, an exploration of hybridity. Her work sold at Christie’s as part of Augmented Intelligence and has gained institutional recognition across major museums.
AI art is selling for millions. But the same market that legitimised AI at seven figures is simultaneously driving a 66% surge in purchases of small-scale handmade paintings. Both trends are accelerating simultaneously.
This is the paradox at the centre of the 2026 art market. AI has not replaced handmade painting. It has, in some measurable way, increased its value. Collectors want objects that cannot be replicated with a prompt. AI solved efficiency but created infinite replicability - the exact opposite of what drives art value.
The artists painting alongside it
Away from the auction rooms and the courtrooms, something quieter is happening in studios across Europe and North America. Painters are incorporating AI into their practice - not as a replacement for the brush, but as a tool for thinking.
A London painter recently used Midjourney to generate 50 variations of a composition, selected one as a structural foundation, then layered hand-painted textures over the AI-generated starting point. A Berlin-based multimedia artist described her process this way: “I use AI to generate the starting point, and then my job is to destroy it, rework it, humanise it.”
Rebecca Taylor, a UK painter, set herself a deliberate challenge at the start of 2025: to use ChatGPT to suggest themes, colour palettes, and mediums, then translate the suggestions into a physical oil painting entirely by hand. The result - Regeneration After Disruption - documented an entire working process that is now shared widely among artists navigating the same question.
Professor Rebecca Xu of Syracuse University, whose animated work has featured in international AI conferences, frames it in historical terms: “When the camera first came along, it shook up the art world the same way AI is shaking up creativity right now. Before the camera, painting was the primary visual art form to capture reality. The camera changed everything. Many traditional painters thought cameras were taking over. But that wasn’t true. The invention of the camera helped artists create new ways to express themselves.”
The analogy is not perfect. The camera did not scrape the portfolios of every living painter without permission to learn how to make images. But the deeper point - that new tools tend to redefine rather than eliminate - has historical weight behind it.
The Artsy AI Survey 2026, based on responses from more than 300 gallery professionals, found that there is still no clear industry definition of AI art: 28% of respondents say they do not have a formal definition at all. 22% define AI art as fully prompt-based or generative works where the primary composition is AI-generated. The market is debating a category it cannot yet define.
Where does your painting sit in a market that is changing this fast? Understanding the value of your work has never been more important - or more possible.
Get your free painting valuation →
What this means for painters pricing their work
The AI disruption has a specific and underappreciated implication for how emerging artists should price and position their paintings.
Handmade is a premium signal, not a baseline assumption. In a market increasingly populated by AI-generated imagery, the fact that a human being painted something - slowly, with physical materials, over real time - is no longer obvious. It needs to be stated. Your artist statement, your process documentation, your social presence, and your pricing all communicate this. Price too low and you signal that handmade craft is fungible with machine output. Price with confidence and you position your work in the category that the 2026 data shows is gaining value, not losing it.
Provenance matters more than ever. AI tools promise efficient art appraisal but cause “slip-ups based on misinformation,” according to The Art Newspaper’s December 2025 market analysis. Collectors face bad provenance checks. In this environment, a painting with clear, documented provenance - where it has been shown, who has owned it, what it sold for - is more valuable relative to undocumented work than it was three years ago. Document everything from day one.
The mid-market for handmade originals is growing. Fine art auction sales hit $4.7 billion in H1 2025, with the $1 million to $10 million band rising nearly 14%. Small-scale paintings under 40 square inches saw a 66% purchase increase in 2025, representing 40% of all art purchases. This is not the AI tier and it is not the trophy tier. It is the range where collectors buy work they live with - and it is the range where most emerging artists operate. The data says this is where the growth is.
AI has not made your paintings worth less. It has, arguably, made the case for their value easier to make - provided you understand that case and price accordingly.
What comes next
The September 2026 trial in Andersen v. Stability AI will be the most consequential legal event in the art world in years. A ruling in favour of the plaintiffs could force AI companies to rebuild their training datasets from licensed works, fundamentally restructuring the economics of generative AI. A ruling against them would confirm that scraping publicly available images for training is legally permissible under current US copyright law - and would likely accelerate both AI development and the political pressure for legislative change.
The European AI Act, fully applicable from August 2026, has already begun imposing transparency obligations on training data. AI companies operating in Europe must now disclose what material was used to train their models - a requirement that changes the dynamic for European artists in ways that are still becoming clear.
The AI art market is expected to grow by nearly 29% annually, reaching over $40 billion by 2033. Approximately 35% of fine art auctions now include AI-created artworks.
The camera did not end painting. It changed what painting was for. The question for 2026 and beyond is not whether AI will change the art world - it already has. The question is who benefits from that change, and on whose terms.
At Priceyourpainting, we use AI to do one specific thing: give emerging artists access to the market data that has always been available to gallerists and dealers, but never to the painters themselves. Our valuation is not AI-generated art. It is AI-assisted knowledge - and there is a difference worth understanding.
See what your painting is worth in the current market →
References
Andersen v. Stability AI: The Landmark Case Unpacking Copyright Risks of AI Image Generators - NYU Journal of Intellectual Property & Entertainment Law, 2024. jipel.law.nyu.edu
AI Copyright Lawsuit Developments in 2025: A Year in Review - Copyright Alliance, January 2026. copyrightalliance.org
The Artsy AI Survey 2026: What Galleries Really Think About AI in the Art World - Artsy, March 2026. artsy.net
From MoMA to Sotheby’s: How Refik Anadol Scaled the AI Art Market - MutualArt, March 2026. mutualart.com
AI Art Market: Anatomy of a $3 Billion Boom - Univile, February 2026. univile.com
AI Art Is Selling for Millions: So Why Are Collectors Buying Handmade Again? - UC Strategies, February 2026. ucstrategies.com
Ways of Seeing, and Selling, AI Art - arXiv, March 2025. arxiv.org
AI Art Trends to Watch in 2026 - Unite.AI, January 2026. unite.ai
How Artists Are Embracing Artificial Intelligence - Syracuse University, August 2025. news.syr.edu
Made with love by artists, for artists.
Stay connected with us on Instagram for more artist stories and market updates.