For the last few months, I’ve been talking to writers, editors, professors, technologists, and publishing people about AI writing detectors and what it means for the entire writing ecosystem. There seems to be a new controversy every week where someone loads an author’s work into @pangram and gleefully posts a score “proving” that the writer’s work is synthetic. It’s a new sort of literary scandal, but one with much fuzzier contours. For centuries the cardinal sin of writing was plagiarism, but plagiarism is something easy to define objectively. You compare two pieces of writing and see how close they are to one another before making a call about whether one author has lifted another’s work. Unless a writer admits to it or agrees to a full forensic computer investigation, figuring out whether they used AI to write for them is more difficult to determine. That doesn’t mean that these writers accused of cheating the system didn’t use AI, it just means that there’s a fog over the accusation that may never lift. In talking to @max_spero_ for this story a while back, we talked a lot about the truth and why the truth is important. I believe Max cares a lot about the truth and his mission in wanting to insulate humans from asymmetrical slop warfare is great. But in reporting this story out I think that, far from being a tool that reveals the truth, AI detectors are injecting something far more chaotic into the conversation about writing: Doubt. Read the whole story for @CNN here: cnn.com/2026/10/21/tech/ai-d…