AI tools can be valuable in the early stages of the writing process as you are working to flesh out ideas, outline your content, and conduct research to support your claims.

Brainstorming Techniques Using AI

Tools that industry experts note for this category of work: Thinking or pro (rather than fast) versions of LLMs

Examples: Gemini Thinking or Gemini Pro; Claude 4.6 or 4.6 Sonnet; OpenAI o4 or ChatGPT-5 Pro

• Talk (or type) through your early ideas with an AI tool like Gemini, Claude, or ChatGPT as you would with a friend who asked you about your work. The conversational style of these AI tools may help to prompt creativity, connections, and extensions of your ideas in much the same way a human brainstorming partner would. Dan Rockmore compellingly describes this process in his New Yorker essay “What It’s Like to Brainstorm with a Bot.”
• Generate an expansive list of ideas using AI, then whittle them down yourself. AI can come up with lots of ideas very quickly. Most of them won’t be great, and many of them won’t fit your voice, perspective, or message. But having a long list of ideas to reject, modify, or embrace is often much easier than staring at a blank page.
• Request and respond to customized writing prompts to help you get unstuck. Use the AI tool to generate questions you can respond to as you draft your manuscript. Try prompts like: “I’m writing a chapter about X in my book about Y. I know I want to say A, B, and C. Provide me with questions to aid me as I’m drafting.” Or “Give me a writing prompt to help me start drafting my chapter about X for my book about Y.”

Guardrail Tip: As you are considering how you will use AI tools to support you in your work, and considering where AI’s work should stop, weigh this question: If the work you’re asking AI to do were being done by a person, would you need to give that person co-author or contributor credit? If so, you may want to reevaluate what tasks you’re outsourcing to AI.

Structuring and Organizing Your Writing with the Help of AI

Tools that industry experts note for this category of work: Source-grounded LLMs that base their responses largely on local documents provided by the user

Examples: NotebookLM, Nouswise, AnythingLLM

• Develop a template for structuring your chapters with an AI tool by telling it what kind of book you’re writing, who your target audience is, and what kind of content you plan to include (for example, let your AI tool know whether you’re planning to include elements like case studies, end-of-chapter key takeaways, or reader self-assessments). You might ask the AI tool to generate a few ideas for chapter templates, which you can then evaluate, combine, or finesse. Give the tool feedback about the templates it generates for you to further refine the output so that it works for you. This is initial work; you’ll want to remain open to receiving thoughtful, next-level input from expert editors and designers. Publishing industry experts can help you to differentiate your book while still fitting in the genre, introduce nuance, expand on your human creativity, and support you in complex ideation.
• Free-write or dictate, then ask AI to suggest structure. It is often easier to draft raw content when you aren’t worried about organizing it or making it flow well. Once you have some raw content down on paper, AI can help you identify common themes, compelling structures, or opportunities to connect ideas. If you’ve already used an AI tool to help you develop a template for your chapters, ask it to suggest where your content might fit into that template. Use caution when asking AI to implement its suggestions; it is likely to modify the content you’ve written without being prompted to do so and without informing you that it has done. Unsolicited AI revisions can impact tone, accuracy, nuance, and overall reader experience. It is more prudent to take the suggestions and implement them yourself.
• Decide on the best format for articulating particular ideas by asking an AI tool to show you options or brainstorm when certain formats are more fitting. For example, if you’re describing a study in your book, you may wonder whether you should communicate the study results in a table, a list, or short narrative paragraphs. AI can generate the different options to help you visualize and explore the decision.

Guardrail Tip: The process of writing is the process of thinking. When we write, we think with our fingers; the act of putting the words on the page prompts you to go deeper, coming up with better insights, allowing you to offer more to your reader than you could if you were only revising content that has been provided to you from elsewhere.

Conducting Research with AI Tools

Tools that industry experts note for this category of work: Tools designed to produce research reports and cite their sources

Examples: Perplexity Deep Research, ChatGPT Deep Research Mode, Consensus

• Refine your understanding of your target audience by prompting AI to explore what your target audience likes, what their pain points are, etc. You likely have a general idea of who your reader is; AI-informed research can help you think through additional details so that you can create a more nuanced picture.
• Understand the competitive field by using AI tools to research books on the market. AI tools can summarize reviews and industry response to competitive titles, provide high-level summaries of book content, and help point you to other authors who are writing in your space. This understanding can inform decisions you make about what to include in your manuscript and how your book might be positioned in the marketplace.
• Conduct research to find external sources to cite in your book using AI tools that are specifically designed to conduct deep and accurate research. As with all AI-supported research, always read the source material and verify your sources independently.

Guardrail Tip: Remember, AI tools gather baseline research, similar to the work that might be performed by an entry-level research assistant. You are the expert. It is your responsibility to confirm, review, and build upon AI’s findings. Don’t take it on faith that AI results are accurate, on-point, or adding the most value possible. That’s your job.

Further Reading

Check out the following articles from Greenleaf’s Learning Center to further explore how AI factors into your writing and publishing journey:

• “How AI Search and Retail Discovery Are Changing Book Marketing”
• “The Pros and Cons (but Mostly Cons) of Using AI to Create Your Book”
• “The Future of Publishing: AI, Hybrid Publishing, and Author Power”

Guardrail Tips

While AI can be a great brainstorming partner, research assistant, and seer of patterns, it is not . . .

• YOU! Remember, your readers are buying your book because the want your voice, expertise, and perspective. Use AI tools to support you in delivering your best work to your readers rather than using it to create work in your stead.
• skilled at synthesis and thinking. AI tools are best suited to perform administrative, foundational, or simple tasks so that you can step in to perform the high-level thinking and creative synthesis of ideas. If you are thinking that AI is going to make the whole process faster, you’re thinking about it the wrong way. Instead, consider that AI is a supportive tool to help you dedicate that same time to higher-level work. AI shouldn’t be used to reduce the total amount of time that you dedicate to the project; rather it frees up time and space in your schedule to dedicate to deep thinking.
• capable of creating fully copyrightable material. Everything that AI generates must be substantially revised or thoughtfully integrated into a larger work in order to be protected by