Can AI Steal Your Book? The Alarming Plagiarism Problem! | US Publishing Expert
In this episode of An Hour of Innovation podcast, host Vit Lyoshin speaks with Julie Trelstad about one of the most urgent and under-discussed challenges in publishing today: how artificial intelligence is changing plagiarism, authorship, and content ownership.
Julie explains how AI makes it possible to copy books at scale, rewriting, repackaging, and republishing them under different titles and author names, often within days of release. These copies can look legitimate to both readers and online marketplaces, allowing them to spread before authors even realize their work has been taken. The conversation highlights why traditional copyright protections struggle in a world where machines, not humans, are the primary consumers of content.
The episode explores why legal frameworks alone are no longer sufficient, and why machine-readable identifiers, metadata, and content registries may be critical to restoring trust and accountability. Julie also discusses what a “post-scraping age” could look like; one where permission, provenance, and attribution are built into the infrastructure of publishing rather than enforced after the damage is done.
This is a thoughtful, grounded discussion for authors, publishers, and anyone interested in AI ethics, intellectual property, and the future of creative work, focused on real risks and practical paths forward.
Julie Trelstad is a publishing executive and strategist known for her work at the intersection of technology and intellectual property. She has spent decades helping publishers, authors, and platforms navigate the identification, protection, and trust of content at scale. In this episode, her perspective matters because she explains not just that AI plagiarism is happening, but why the system makes it so hard to detect and stop, and what could actually help.
Takeaways
- AI can clone and resell a book in days, and most platforms struggle to reliably prove that the theft occurred.
- AI-generated plagiarism often looks legitimate enough to fool retailers, reviewers, and buyers.
- Authors lose sales and reputation when fake AI versions of their books appear at lower prices.
- Traditional copyright law exists, but it was never designed for machine-scale copying and AI training.
- There has been no machine-readable way for AI systems to recognize who owns content, until now.
- Content fingerprinting can detect similarity across languages and paraphrased AI rewrites.
- Time-stamped content registries can establish legal proof of who published first.
- Most books already inside AI models were scraped without the author's consent or compensation.
- AI lawsuits focus less on training itself and more on the use of pirated content.
- Authors could earn micro-payments when AI systems use specific paragraphs or ideas from their work.
Timestamps
00:00 Introduction
01:37 Why AI Plagiarism Is So Hard to Detect
03:25 Amlet.ai and the Fight for Content Ownership
05:32 How Copyright Worked Before Generative AI
08:09 The Origin Story Behind Amlet.ai
12:22 Building Machine-Readable Infrastructure for Copyright
14:24 How Publishing Is Changing in the AI Era
17:34 How Authors Can Protect Their Work with Amlet.ai
20:38 Tools Publishers Use to Detect and Enforce Rights
21:38 How Authors Can Monetize Content Through AI
24:27 The Reality of AI Scraping and Plagiarism Today
27:00 Publisher Rights, Digital Security, and Enforcement
29:08 Evolving the Business Model for AI Licensing
35:34 The Future of Digital Ownership and AI Rights
38:37 Innovation Q&A
Support This Podcast
- To support our work, please check out our sponsors and get discounts: https://www.anhourofinnovation.com/sponsors/
Connect with Julie
- Website: https://paperbacksandpixels.com/
- LinkedIn: https://www.linkedin.com/in/julietrelstad/
- Amlet AI: https://amlet.ai/
Connect with Vit
- Substuck: https://substack.com/@vitlyoshin
- LinkedIn: https://www.linkedin.com/in/vit-lyoshin/
- X: https://x.com/vitlyoshin
- Website: https://vitlyoshin.com/contact/
Episode References
International Standard Content Code (ISCC)
https://iscc.codes
An international standard that generates digital fingerprints for content, allowing similarity detection across formats and languages.
StreetLib
https://www.streetlib.com
An international ebook distribution platform that helps authors publish globally and integrates content identifiers for ownership protection.
Amazon Kindle Direct Publishing (KDP)
https://kdp.amazon.com
Amazon’s self-publishing platform, referenced in the context of reporting plagiarism and filing takedown complaints.
Amazon
https://www.amazon.com
Cited as a major marketplace where AI-derived or plagiarized books can appear quickly and look legitimate.
Barnes & Noble
https://www.barnesandnoble.com
Mentioned as a retailer actively monitoring and responding to plagiarized or infringing book listings.
Kobo
https://www.kobo.com
A global ebook retailer referenced as part of the ecosystem dealing with AI-related book plagiarism.
robots.txt
https://developers.google.com/search/docs/crawling-indexing/robots/intro
A web standard used to tell automated systems what content they can or cannot crawl, cited as an analogy for missing tools in book publishing.
Digital Rights Management (DRM)
https://en.wikipedia.org/wiki/Digital_rights_management
Encryption-based systems historically used to restrict ebook copying, described as ineffective against large-scale piracy.
Google Books
https://books.google.com
Referenced in discussion of earlier copyright lawsuits and how large-scale digitization eventually became a discovery tool rather than a threat.
Anthropic
https://www.anthropic.com
Mentioned in the context of U.S. lawsuits addressing whether AI companies can legally train models on copyrighted books.
OpenAI
https://www.openai.com
Referenced alongside Anthropic regarding legal rulings that distinguish purchased content from pirated training data.
The Authors Guild
https://authorsguild.org
A professional organization for writers that challenged Amazon’s Kindle chatbot features through legal action.
Napster
https://en.wikipedia.org/wiki/Napster
Referenced as a historical parallel to early digital music piracy and the industry’s struggle with licensing.
iTunes / Apple Music
https://www.apple.com/apple-music/
Used as an analogy for how licensing infrastructure eventually stabilized digital music distribution.
Spotify
https://www.spotify.com
Mentioned as a comparison point for potential future micro-payments when AI systems use licensed content.
