Most authors are uploading their book manuscripts into ChatGPT, Claude, Gemini, Perplexity, and other AI tools without any idea what actually happens to their intellectual property when they hit enter.

In this video, business and intellectual property attorney Jessica Eaves Mathews of Leverage Legal Group walks authors through exactly what is safe, what is dangerous, and what most writers get catastrophically wrong.

Learn why your copyright survives the upload but why that is not the risk that should concern you, why toggling training off does not equal privacy, and how the New York Times lawsuit against OpenAI forced the company to preserve chat histories that should have been deleted, exposing millions of users’ prompts to potential third party access.

Jessica walks through the exact terms of service language authors should look for, why free AI tools should be treated as if the author is the product, how enterprise and offline tools such as LM Studio and local Llama and Mistral installations offer the highest levels of protection, and why authors under publishing contracts may be breaching confidentiality or exclusivity clauses the moment they paste their manuscript into a prompt.

She also explains the four step manuscript protection protocol, the categories of content that should never touch cloud based AI, and the moving target problem of platforms silently changing privacy policies to sweep more user content into their training sets.

Watch to the end for a direct offer to have your AI use, your content strategy, and your publishing agreements reviewed before your next manuscript becomes someone else’s training data.