Hyperallergic Editor to Art World: Stop Using AI to Write

A new opinion piece from Hyperallergic opens with a paragraph that its author says was generated by Claude, the AI chatbot, in response to a prompt for an introduction to an essay against AI use in art writing. It reads: “Wall labels. Press releases. Artist statements. In today’s fast-paced art world, the written word stands at a crucial crossroads.” The point, as the piece makes clear, is that readers who recoil at that cadence may be among those who have developed what the author calls a strong aversion to generative text.

The essay, written by an editor at the publication, argues that the telltale style is easy to spot: overconfident in tone, hollow in substance, like a motivational LinkedIn post that feels relatable at first and leaves little behind. People who read, write, and edit for a living, the author suggests, are especially attuned to what the piece describes as the artificial staccato and mawkish pathos of machine-made prose. Still, the author writes, its spread through the museums, galleries, and fairs of the so-called art world is undeniable. As evidence, the editor recounts a friend, an artist, who sent along an obviously AI-drafted press release from a well-respected Manhattan institution, with the comment: “Wow, they really have no shame.”

The essay is notably sympathetic to individual artists. Describing itself as sensitive to how time- and resource-intensive writing can be, it points to the burden of self-promotion that has become untenable for many, particularly those without a gallery or an assistant. Artists are drafting exhibition texts, lengthy grant applications, and social media copy for the online presence now expected of them, the author notes, often with little training and no compensation. When tools emerge to optimize those chores, the piece argues, creative people will reach for them not because they undervalue writers, but because they want to protect their own artistic labor.

The verdict is far less forgiving for institutions. Well-resourced museums, foundations, publications, and other cultural organizations, the author contends, have no excuse for relying on AI for textual materials. The same goes, according to the essay, for PR representatives, marketing professionals, and anyone whose principal job is to communicate adeptly with the public and the press. The author adds a caveat aimed at employers: underpaid, overworked museum workers must be given more time and resources, so that delegating original writing to AI does not become a temptation.

Journalists and other writers, the piece continues, face similar precarity to artists but carry a moral responsibility to keep AI out of the heart of the writing process, which it defines as crafting original sentences whose form deepens their substance. The author cites the guidelines of Wired, which state that someone who writes for a living needs to constantly be thinking about the best way to express complex ideas in their own words.

The essay lands in New York at a moment when the city’s cultural institutions are heading into a busy autumn season of openings, fairs, and fundraising, each of which runs on a steady stream of press releases, wall texts, and grant language. Hyperallergic’s argument, in short: if an institution can afford to pay a writer, it should.