The conversation concerning A.I.’s role in human creativity has gained interest over the last couple years as its presence has grown across artistic mediums and digital platforms, and is now expanding geographically with the springing up of data centers across the U.S. While the top executives at multimillion-dollar organizations – including the Grammys, the Oscars, and Vogue – have embraced the use of A.I. to freshen up their multimillion dollar projects, such as “The Brutalist” starring an A.I-generated Hungarian accent, the question of A.I.’s utility in creative processes and the creative industry remains, despite what corporate execs want you to think.
There are several discourses to be had between artists in their respective mediums about what a “proper” use of A.I. might be when creating something, but here I want to focus on A.I. use in writing – whether it’s essays, articles, or emails. I will also elaborate on why I believe A.I. as a technology cannot be treated or considered the same as previous technologies that have facilitated human creativity, and why transparency about A.I.-use in writing is absolutely necessary to differentiate what humans independently write, and what they write with the help of an LLM (large language model). This is a terribly pathetic conversation to have in 2026 considering we’ve never had to deal with this before.
First, some clarification. I will not be addressing writing that has been completely generated by an LLM; that is not human writing, and I don’t care how good the input prompt is written. If you want to be a writer, you need to stroke the keys of the keyboard onto a document, not into a textbox. If you’re researching like hell to write a good prompt, then you should be able to write that researched response yourself. Sorry not sorry to the ‘friends’ of A.I. who consider themselves writers but are also interested in taking massive shortcuts — you will not be considered in this conversation.
We do, however, have to confront the issue of A.I.’s proliferation of human communication. I firmly believe that in today's digital communications, A.I festers in several corners of the digital world and is seeping into our everyday lives from pervasive exposure to its writing. Unfortunately, execs like Sam Altman – CEO of OpenAI – were right when they said A.I. would be everywhere. Today it is built into most of our search engines and apps, and an entire market obsessed with “productivity” and “convenience” has sprung up to do the work of summarizing our resources and notes. People have started falling into the convenience trap of having their works polished by A.I.: either by assisting with research, or acting as an editor. As a result, many articles and essays written and disseminated on the internet today have high chances of being touched by an A.I. writing tool. Back in September, the Daily Californian – UC Berkeley’s student newspaper – released an article explaining how they ran over 1,000 city and campus correspondence records through A.I detection software and found its writing across leadership, legislation, and public announcement documents. Of the resolutions passed by the UC student government, 42% were flagged for A.I. writing. I’m willing to bet this is not just happening in Berkeley. It’s happening at institutions, educational and otherwise, all across the country.
This hunch isn’t intended to be fearmongering. Instead, I think the Daily Cal’s article proves two things: 1) that widespread use of A.I. writing tools in our daily communications across the web is increasingly common, so much of its writing is available everywhere, and 2) A.I. indeed exhibits a detectable style of writing, one which is separate from organic, human writing. This expansive use even reaches the writers that try their hardest to keep it away, and everyone else in the process. As any writer should know, we are influenced by what we read, meaning A.I.-generated text has started to influence many of us, whether we notice it or not. It sounds grim, inescapable, but on the bright side, we’ve dealt with this techno-nightmare long enough to know that we can detect A.I.'s writing ourselves—because it sounds like slop.
So, where does that leave writers who use A.I. writing tools? We are well aware that A.I. is an imperfect technology, and using it to do research or review requires having to still verify what it tells you to maintain the integrity of your work. I would argue this process takes just as long as doing the research yourself, and you avoid having A.I. rifle through documents only for the information to be wrong. What about using A.I. as an editor? I believe the constant use of A.I. editors eventually make you sound like A.I., so writers should be careful about consulting an LLM to constantly “refine” their work. That’s what real editors are for, and if you’re making A.I. rephrase your own words (multiple times over), are you really the one articulating your thoughts? I say no, and the more a writer uses an A.I. to edit their work, the closer they inch towards sounding like a machine, to having A.I. tell them what to say.
Naturally, the creatives who embrace A.I. in their process — not just my fellow writers, all mediums — have created their own arguments for why it is not an obstacle to creation, to creativity. That it is — like all other technologies humans have invented to aid in creating—the latest tool at humanity’s disposal to realize itself in art. With this I vehemently disagree. Artificial intelligence is not like other technologies we’ve created to facilitate expression. As a Marxist, I adhere to the notion that humanity’s ideas have changed with the modes of production; the printing press gave us Shakespeare, and Adobe photoshop made the computer an art studio. A.I. is not photoshop and it is not the printing press – it's meant to sever creativity from us altogether. Why do I make these remarks? Look at how A.I. harms our physical and cognitive faculties. A 2025 study at MIT observed the potential cognitive effects of regularly using chatbots to write essays. In the study, subjects wrote SAT-style essays with the help of ChatGPT, the Google search engine, or nothing at all. According to the study, ChatGPT users “consistently underperformed at neural, linguistic, and behavioral levels.” Our kids will never be able to develop critical thinking if they constantly have access to an LLM that will take it from them. That is exactly what the primary author of this study, Nataliya Kosmyna, is worried about, and what I am worried about too. Studies like Kosmyna’s only provide further evidence for how harmful this technology is to human thought and creativity. It’s cutting it at the root.
One last point I want to reach about A.I. is its role in our broader capitalist society. A.I., to me, is the highest form of the productive worker, an idealized form forced to give arm, limb, and labor to grease the wheels of productivity. This worker has no arms and limbs to organize when conditions are shitty or when wages are pitiful. With A.I. there are no conditions to maintain or wages to pay. Earlier I mentioned that those who utilize A.I. editors try to bypass what real human editors do. Should these editors’ labor be handed off to a machine too? Like all the other labor A.I. has already absorbed from the workforce? This technology is recursive, it is self-generating, it scrapes its own prior data for generation. It's beginning to exhibit signs of awareness and autonomy, like top A.I. officials are warning, and it will make stronger, smarter versions of itself. Last I remember, the printing press never worked independently from humans, it never made better printing presses that could supposedly write better books. Capitalism breeds alienation from everything, and with A.I., we’ve opened the door for us to be alienated from our own thoughts. That is fascism. That is why the billionaires, the owners, the investors, the bourgeoisie, want it so badly.
So what is to be done? Well, for one, we need to be more conscious of the fact that writers are regularly engaging, directly and indirectly, with A.I. writing tools and its generated text in general. We also must acknowledge that A.I. writing is making its way into our vernacular, and we should continue to have conversations with fellow writers about how we use, or become influenced by A.I. if we’re interested in continuing to release works to the public. If we meet writers that consistently use A.I. as an editor, that must be explicitly communicated to the other writers, editors, as well as readers. The public especially deserves to know when they are reading A.I.-assisted work. What I call for is transparency about A.I.-use in writing, because there is a very real, and very present, threat to human writing when we let A.I. articulate something “better.” I write this essay as a writer and reader, and believe that writers must communicate when A.I. enters their workflow in any capacity in order to preserve human thought, creativity, and ideas. If you don’t, those who don’t use A.I. will pick up on it eventually.
This essay was written and edited by members of The Berkeley Pulse and Berkeley Barb, and not edited or assisted with A.I.