AI Best Practices for Authors: A Comprehensive Guide

Introduction to AI in Writing

I am an author who is curious about how to use AI ethically in my writing. Can I use AI at all, and if so, how do I need to disclose that use?

Yes, you can use AI as a tool in your writing process, but you must disclose that use to your publisher and your readers if you used AI for anything more than brainstorming or editing. Disclosing AI use is an important part of maintaining ethical standards in writing, especially in light of the potential for AI misuse. Keep reading to learn about specific publisher and platform policies regarding AI use and to learn more about how to disclose AI use in your manuscripts.

AI Best Practices for Authors

The use of AI in writing is rapidly increasing, presenting both exciting opportunities and complex challenges for authors. AI writing tools, such as ChatGPT and Grammarly, can assist with tasks like brainstorming, research, outlining, and editing. While these tools can offer valuable support, it’s crucial for authors to use them ethically and transparently. This guide provides a comprehensive overview of AI best practices for authors, addressing key considerations regarding ethics, disclosure, specific use cases, legal aspects, and future trends.

Ethical Considerations

• AI’s Impact on the Writing Profession: The rise of AI raises concerns about potential job displacement and the need for writers to adapt their skills. Authors must stay informed about these developments and proactively develop their expertise to thrive in an evolving landscape.
• AI as a Tool, Not a Replacement: AI tools should be used to augment human creativity, not replace it. The goal is to use AI to support the writing process while maintaining the unique voice and originality that define human authorship.
• Originality and Authorship: Authors must ensure that they are the true authors of their work, even when using AI assistance. Rewriting and revising AI-generated text in your own voice is essential to preserve originality and maintain authorial control.
• Respecting Copyright and Avoiding Plagiarism: Using AI to copy or mimic the style of other writers is unethical and could lead to accusations of plagiarism or copyright infringement. Thoroughly fact-check all AI-generated content and ensure that all sources are properly cited.
• Solidarity with Other Creatives: Authors should be mindful of the impact of AI on other creative professions, such as illustrators, voice actors, and translators. Advocating for ethical AI practices and supporting human creatives in these fields is crucial.

Disclosing AI Usage

Transparency is paramount when using AI in writing. Disclosing AI usage is essential for maintaining trust and credibility with readers and publishers, ensuring ethical practices, and complying with platform and publisher policies.

Reasons for Disclosure

• Transparency and Reproducibility: Disclosure allows other researchers to understand the methods used in a study, which promotes transparency and makes it possible to replicate and verify research findings.
• Ethical Considerations: Openly acknowledging AI use addresses ethical concerns related to authorship, plagiarism, and the potential biases inherent in AI systems.
• Trust and Credibility: Disclosure builds trust with readers and publishers by demonstrating honesty and a commitment to responsible AI practices.
• Compliance: Many platforms (e.g., Amazon KDP, Medium) and publishers have specific policies requiring disclosure of AI usage. Failing to disclose could result in consequences such as restricted distribution or rejection of the work.

To Whom to Disclose

• Publishers: Most publishing contracts contain clauses that guarantee the originality of the author’s work. Using AI-generated content without disclosing it to the publisher could be considered a breach of contract. Publishers also need to be aware of the use of any non-copyrightable material, such as AI-generated text, to ensure proper copyright registration and protection of the work.
• Readers: Ethically, readers have a right to know if the content they are consuming has been generated with the assistance of AI. There are various ways to disclose AI use to readers, such as including an author’s note, a statement in the acknowledgments section, or a disclaimer at the beginning or end of the work.
• Platforms: Platforms where you publish your work, such as Amazon KDP or Medium, may have specific policies regarding AI usage disclosure. Check the platform’s terms of service to ensure compliance.

How to Disclose AI Usage

There are three main types of signals authors can use to disclose AI usage: behavioral, verbal, and technical.

Behavioral Signals

The presentation of the content itself can signal AI involvement. For example, using an avatar for a chatbot or choosing a synthetic voice for an audiobook can indicate that AI is being used. It’s important to avoid giving bots overly human-like attributes to avoid misleading the audience. “Creator fidelity” is key: be transparent about the fact that the content was created with the assistance of a bot.

Verbal Signals

This involves using explicit statements to inform the audience about AI usage. Instead of listing AI as an author, which is generally considered unethical, focus on explaining how AI was used in the creation of the content. Several structures can be used for verbal disclosure:

• Content labels: Short tags that indicate the content was generated by AI.
• Visible watermarks: Similar to watermarks that indicate “draft” or “sample” content, watermarks can also be used to signal AI generation.
• Bylines: When AI is involved in content creation, it’s important to carefully consider how bylines are handled. If an individual has substantially rewritten AI-generated output, they can be credited in the byline. However, if the bot was primarily responsible for the content, using credits like “produced by” or “reviewed by” may be more appropriate.
• Disclosure fields: These provide additional information about how AI was used, such as disclaimers or warnings.
• Acknowledgments: This section can be used to provide more detailed information about the AI tools used and their specific contributions to the work.

Technical Signals

These are machine-readable statements embedded within the content that can be used to verify its authenticity and origin. Examples include:

• Metadata: This data about the content, such as the date it was created and the tools used to create it, can be embedded in the file.
• Technical watermarking: This invisible watermark can be used to verify whether the content has been altered.
• Cryptographic signatures: These digital certificates can be used to verify the creator of the content.

Specific Use Cases and Examples

How AI is used and disclosed will vary depending on the type of writing.

Fiction Writing

AI can be used for brainstorming plot ideas, developing characters, and even generating dialogue or scenes. It’s essential to ensure that the AI output is thoroughly rewritten and infused with the author’s unique style to avoid relying on generic or predictable content. An example disclosure for fiction could be: “This novel was written with the assistance of AI for brainstorming and plot development. However, all final writing and creative decisions were made by the author.”

Non-Fiction Writing

AI can be a valuable tool for research, outlining, fact-checking, and editing. Authors must ensure that any AI-generated content is accurate, properly cited, and integrated seamlessly into their own writing. Example disclosures for non-fiction could include:

• Article: “This article was written with the assistance of AI for research and fact-checking. All analysis and conclusions are those of the author.”
• Blog Post: “This blog post was edited using AI for grammar and clarity. The content and opinions expressed are solely those of the author.”
• Book: “This book was written with the assistance of AI for outlining and research. However, the author has thoroughly reviewed and revised all content to ensure accuracy and originality.”

Academic Writing

Many universities and academic journals have strict policies regarding AI usage. Authors must carefully review these policies and disclose AI usage accordingly. Some institutions may require students to provide detailed information about the AI tools used, the prompts entered, and the extent to which the AI output was modified. Example disclosures for academic writing could include:

• “AI Usage Disclosure: This document was created with assistance from [specify the AI tool]. The content can be viewed here [add link] and has been reviewed and edited by [author’s full name]. For more information