France just handed its biggest literary prize to a guy who put ChatGPT in the book on purpose.
Tuesday night in Paris, Philippe Jaenada won the 2026 Prix Goncourt for L’Inconnue du quai de Javel (Flammarion)—a long dive into the unsolved 1949 murder of a young Bretonne by the Seine. Next morning on France Inter he said he’d used AI “three times.” Once to hide a ChatGPT sentence in the novel so he could play with the reader (he told the Goncourt jurors which line it was; he’s not telling the rest of us). Once to mock ChatGPT’s absurd answers to a technical question. Once for research. The exchanges with ChatGPT and Claude are already in the book, pages 237–240. He wasn’t caught. He wrote it down.
Le Parisien ran the whole novel through Pangram. The detector said zero words were machine-generated. It missed the sentence he planted. Surprise.
Same prize season: Thélyson Orélien got pulled from the shortlist over AI suspicions. The Académie Goncourt’s president wants the rules rewritten by year-end so the prize honors “a work written by a human being.” So the academy is drafting a formula to ban AI-aided books… while crowning a novelist who used the tools out loud, as material, and as a joke. Jaenada’s line is the useful one: you can’t pretend it doesn’t exist; use it, name it, don’t let it replace you.
That’s the split that matters. Not “was there a detector score.” Not a three-tier label on the spine. The split is between someone who put the machine in the scene and someone who tried to disappear behind it.
Circle is being built for the first kind.
Members request assignments. Reviews are structured and prompted, submitted on the site—lenses that push you past “nice” and into what you actually felt. Writers revise and resubmit. Earned credit can go toward your own work or someone else’s. We are not trying to make the law, and we are not drawing up complex formulas for what’s allowed. Before your first submission, you’re asked once to be honest about if and how you use AI. It’s in the membership agreement. Honesty. Your word and your reputation. How you feel about AI writing or production is yours. The one place AI is not allowed is reviewing someone else’s work. We are not interested in how AI feels about it. You can get that from a chatbot. What we want from a review is honest, informed human feedback.
And yes—Circle is new. Almost no members yet beyond the test seeds. We’re recruiting founding members who want to get to work, who’ll have real say in how this grows. The bones are in place. We need builders, not a crowd.
If that’s you, join us and earn your satchel.
Registration is open at gapoz.com. No fees. Membership to the Circle is limited and subject to approval. Satchels are earned, not given away. Come meet people, make some friends, and grow your craft together.
Your name. Your word. Your work.
Also on the beat
- YouTube is testing a 30-day lock for some channels facing spam termination—fix, delete-all, or appeal—while its spam policy still bans AI mass-production of near-identical videos (Tubefilter, Nov 3).
- BookBub’s 2025 survey of 1,229 authors: 45% used generative AI; 74% of those didn’t tell readers. A Mississippi AI officer told the same piece detectors can’t settle authorship (RHCJC News, Nov 4).
- Le Parisien’s Pangram check on Jaenada’s Goncourt winner found “no AI”—and missed the ChatGPT line he admitted planting (Nov 4).
- NormDetect preprint coverage: at a 0.1% false-positive rate, calibrated detectors still miss about half of LLM text (Agora Intelligence on the Oct 30 arXiv paper, Nov 3).
- New arXiv work pushes four-class “human / low / high / fully AI” co-authorship detection and shows paraphrasing attacks wreck detectors trained only on clean data (announced Nov 3).
How this was made
AI assisted research and a first draft of this Circle View. A human reviewed, edited, and is accountable for the judgment and the final take.
- A score is not a confession → Page 1 Still think a percentage is proof? Read Do AI detectors work on fiction?
- A Detector is NOT a Judge → Page 1 The short version of why a score can’t judge a book: Do AI detectors work on fiction?
- Prove you didn’t use a machine → Page 1 If someone’s asking you to prove a negative, start here: Do AI detectors work on fiction?
- Disclosure that never faces the reader → Page 2 Here’s how we keep it simple in Circle, no formulas: How Circle handles AI
- How did you use AI? → Page 2 We ask once, and your word does the rest: How Circle handles AI
- Put it on the Imprint Page → Page 2 Why I’d rather skip the label tiers altogether: How Circle handles AI
- Watermarks don’t sign your name → Page 1 Watermarks, detectors, same problem: Do AI detectors work on fiction?
- They knew the shelf would get crowded → Page 2 Slop is the problem, not the tool. Here’s where Circle stands: How Circle handles AI
- Label the voice → Page 2 How we handle AI in Circle, without the labels: How Circle handles AI
- Byline is no Costume → Page 2 Your name, your word, your work. That’s the deal in Circle: How Circle handles AI
