France has launched a literary prize for books written with large language models, requiring entrants to disclose the AI model and prompts they used. The Prix Pascal puts a practical question about authorship on the page: is declaring AI’s role enough when the work itself is made with its help?
Watch Desk analysis
What happened
The new Prix Pascal de Littérature IA is open to unpublished fiction and non-fiction manuscripts in French of about 300,000 characters, written with the help of large language models such as ChatGPT, Claude or Gemini. Entrants must submit a “seed”, manuscript and technical file together, and declare the model and prompts used. The Hindu reports applications close on 30 November, ten finalists will be announced on 4 December, and the winner will be revealed on 15 December.
The prize’s manifesto says the work should help address an ecological, social or ethical challenge. The Hindu’s K. Srilata, a writer and creative-writing teacher, argues that this framing does not resolve her concerns about AI-assisted writing, including what happens to writers’ work when it is used to train models. Those are her views, not settled findings about every model or writing process.
Why it matters
This is more than a prize with a timely label. Its entry rules make AI use part of the submission record, giving judges and readers a stated account of which model and prompts helped produce a book. That could make the process easier to discuss, though disclosure alone cannot answer who deserves credit or whether training-data permissions are fair.
The debate is also about what writing is for. Srilata describes the work of drafting and revising as part of how writers think, not merely a slow route to a finished text. That is a serious argument, even if readers may draw the line between assistance and authorship in different places.
Our read
The prize has done something useful before a winner is even named: it has made the terms of AI-assisted writing visible enough to argue over. Requiring a prompt record is a start, not a verdict. A technical file can tell readers what tools were used; it cannot, by itself, settle questions of originality, credit or the value of the human work involved.
What to watch
- Whether the rules clarify how judges will weigh the extent of AI assistance.
- What the finalists disclose about their models and prompts.
- Whether the prize’s ethical ambitions include meaningful answers on training data and authors’ rights.
Discussion spark: Should literary prizes accept AI-assisted books if authors disclose their tools and prompts, or does that disclosure leave the central question of authorship unresolved?
Sources and evidence
- Prix Pascal: If AI writes your book, are you still a writer? – The Hindu (11 November 2026, 14:54 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.