AI

Why AI-use allegations against authors may fade: 2026’s scandals stem from books acquired before the growing use of detection tools like Pangram.

Evidence of AI use in early versions of his manuscript has caused Les Éditions du Boréal to drop Haitian-Canadian author Thélyson Orélien.

The manuscript must be submitted alongside an initial prompt, with detail on the language model and generation process used.

The Guild is not demanding publishers abstain from using AI detection tools, but they must be used transparently and with the author’s permission.

After publisher outcry, Meta modified its commercial terms to remove the word train from permitted uses.

One tidbit sailing under the radar is that publishers say the number of authors using LLMs to write their books is much higher than reported.

Starting with Audible Original titles, US and UK readers can have an interactive experience with a character outside of a book’s narrative.

My insights from the AI panel at last week’s Association of American Literary Agents’ People of Publishing conference in New York City.

Humans are checking every manuscript and, in some instances, the Copyright Office sends a follow-up email asking the author to clarify their use of AI.

Publishers report increased AI usage internally for marketing, metadata and title optimization, and admin work.

The Guild says lack of clarity over rights reversion is the most common problem for claims due to lack of publisher records.

Every high-visibility award is at risk of finding itself in this situation if it intends to exclude AI-assisted work from consideration.

The accuracy and funding of research claiming Gen AI causes market harm have become important questions in the consolidated lawsuits.

OpenAI, like Anthropic, created an internal library of books from the pirate website LibGen and even shared those works with outsiders.

Novelist Vauhini Vara has done her own informal testing on Pangram and found she could not trick the system.

Their policy is a strong model but also one of the strictest I’ve seen, banning generative AI for brainstorming or an editing pass.

A new app called iLands allows the creation of autonomous AI agents that not only write articles but pitch their work to media outlets.

GEO is generative engine optimization: trying to influence whether your name or product appears in AI overviews or in chatbot conversations.

Conference attendees have many questions about AI, and organizers best serve their audience by engaging with intent and clarity.

Jane answers writers’ questions about AI detection, disclosure, proof of authorship, and publishers’ AI use, identifying where real risks lie, though firm answers remain elusive.

I divide disagreements into three categories: basic disagreements, surprise agent claims, and even messier disagreements. Reasons for these disagreements vary, but metadata problems and communication from the settlement administrators have led to some of the confusion.

The second annual survey of writers and editors, updated to reflect new AI tools and challenges, is underway.

Ember Novels shows that AI is increasingly able to develop and carry plot through 300 pages with fewer continuity errors than many AI-generated or assisted books currently on the market.

Readers can also generate infographics, audio overviews, quizzes, and other interactive features.