Two stories from the literary world this week are forcing a rethinking of relationship between writers and readers.

On October 11, Vauhini Vara wrote a piece for the Atlantic arguing that the prestigious Caine Prize had been awarded to an author who used AI to write his short story, The Blue Fighter. In an interview with the BBC the story’s author, Simbarashe Steyn Kundizeza, flatly denies the charge and asks why the tool used to flag his story as AI written — Pangram — isn’t being questioned itself. Vara pointed to a dramatic change in style to make her case.

In response Kundizeza wrote: “To suggest that a writer cannot write in different styles or genres, or improve his writing substantially after five or more years fundamentally misunderstands the craft of writing.”

The controversy comes on the heels of Thélyson Orélien, a Haitian-Canadien, who was accused of plagiarism and using AI to write his first book. The New York Times reported on Tuesday that Orélien admitted to plagiarism but denies using AI.

The Orelien case is particularly interesting because it juxtaposes plagiarism and using AI to write. In truth, readers have long had to trust that writers weren’t cheating by plagiarizing work. That trust was built in part because plagiarism was once relatively easy to identify. Simply match it to published material.

There were defenses for plagiarizing, such as that Stephen Ambrose offered when two of his books were flagged for lifting sentences without proper citation. Whether those explanations was enough to satisfy readers was up to the reader.

AI creates another level of challenge. Primarily because AI identification programs themselves aren’t foolproof. A point Ron Charles had great fun with earlier this year in a column about companies awarding certifications for books it claimed to have scanned and certified as AI free. Charles showed how easily that system could be dupped.

Because the tools are unreliable, being charged blames the writer, who is essentially helpless to defend their work, leaving readers to wonder what to believe.

And here is where the mistrust comes in.

When plagiarism was the only worry, readers could approach a work on face value until conclusive evidence proved otherwise.

With AI, readers increasingly must default to an air of suspicion. If award winners are being popped for AI, who can I trust?

It’s a Brave New World readers and writers inhabit. Trust is the key to navigating it.