AI and the future of human art

2026 has been a strange, vertiginous year for anyone who cares about creativity. Scientists published the most complete wiring map of a brain ever made — the male fruit-fly connectome, roughly 166,000 neurons and about 125 million connections, a collaboration between Google Research and Janelia, in Cell. In the same stretch of months, AI systems began generating entire playable worlds from a single prompt, from an AI-rendered Minecraft to Google DeepMind’s Genie 3. We can now read a biological mind and write synthetic ones — and the people building these systems are openly nervous about the pace. Even Demis Hassabis, who leads Google DeepMind, has called for research into AI’s risks to be done “urgently”.

Most of the AI-risk conversation runs to the dramatic — extinction, loss of control, superintelligence. Those debates matter. But there is a quieter harm already unfolding in studios and on kitchen tables around the world, and it rarely makes the headlines: what happens to the people who make art for a living.

The income squeeze is already here

This is not a forecast. It is a measurement. In the Society of Authors’ 2024 survey of nearly 800 creative professionals, 26% of illustrators and 36% of translators said they had already lost work to generative AI, and 37% of illustrators reported their income falling because of it. More than three-quarters of illustrators expected AI to shrink their future earnings. For musicians, an industry study commissioned by CISAC estimated that roughly a quarter of creators’ revenue could be at risk to generative AI by 2028.

Researchers have started to describe the mechanism plainly. A Stanford Graduate School of Business analysis put it in its own title: when AI-generated art enters the market, consumers win and artists lose. Cheaper, faster, infinite supply pushes down the price of the exact commercial work — spot illustrations, stock images, background music — that has long paid the rent while artists made the work that mattered to them.

None of this happens in a vacuum. The International Monetary Fund estimates that around 40% of jobs worldwide are exposed to AI, and the World Economic Forum’s 2025 outlook projects 92 million roles displaced and 170 million created by 2030 — a net gain, but a wrenching reshuffle for the people caught in the churn. “Exposed” is not the same as “erased.” Still, for a working illustrator, the difference is more than semantic.

The scarce thing in an age of infinite images is not another image. It is a person who made one on purpose.

The copyright question no one has answered

There is a second, deeper grievance underneath the income numbers: many of these models were trained on artists’ own work, scraped without consent, credit or payment. Artists are fighting that in court, and the outcomes are still open. In the United States, Andersen v. Stability AI has seen a judge allow the artists’ core copyright-infringement claims to proceed. The New York Times’ case against OpenAI and Microsoft has advanced past early dismissal. In the UK, Getty Images won on trademark grounds — though it narrowed its copyright-training claims mid-trial, so it is not the sweeping victory some headlines suggested.

Taken together — and there are now more than fifty such suits being tracked — the law is unsettled but the direction of travel is clear: courts are treating “we trained on it” as a claim that has to be answered, not waved away. For artists, that is the difference between being raw material and being rights-holders.

Why the human hand matters more, not less

It would be easy to end on despair. We won’t, because the honest picture is more interesting than that. When any machine can produce a competent image on demand, competence stops being the scarce thing. What becomes scarce — and, we’d argue, more valuable — is authorship: a real person, a real hand, a real place and tradition, a real reason this piece exists. A Madhubani painting carries generations of Mithila. A hand-thrown pot carries the potter’s afternoon. Provenance, the plain fact that a human being made this on purpose, is exactly what an averaged model cannot honestly claim.

This is not nostalgia. It is a market signal. The same years that gave us infinite synthetic media have also produced a hunger for the demonstrably human — vinyl records outselling CDs, film cameras back on shelves, people paying a premium for the handmade. When the digital becomes frictionless and forgeable, the analog and the authentic become the luxury.

What platforms owe artists now

If provenance is the value, then the job of a platform is to protect and surface it. That means putting a real person at the centre of every page, keeping the relationship between maker and buyer direct rather than skimming it, and refusing to blur the line between human-made and machine-made. It means treating an artist’s work as theirs.

That is the whole reason KalaCUBE exists. It grew out of Musée Art Café in Dehradun, where local art lived on physical walls; when the world moved online, we took those walls with us. Every profile in our directory of artists belongs to a real person, every piece in the gallery is made by human hands, and artists keep their own relationship with the people who buy their work. AI will keep getting better at making images. Our bet is on the people who make art — and on the growing number of people who specifically want the human kind.