If you spent any time at the London Book Fair this year, you’ll have noticed a pattern. Walk past any panel with “AI” in the title and you’ll find a room full of nervous energy — publishers wringing their hands, authors muttering about theft, and a general sense that the robots are coming for the written word.
Here’s the uncomfortable truth: most of that conversation completely missed the point.
Not because the concerns are baseless. They’re not. But because parts of the publishing industry — an industry that has too often under-served the authors who create its value— has suddenly discovered outrage about rights, ownership, and fair compensation, and pointed almost all of it at the wrong target.
If you’re an established business owner, consultant, or thought leader using a book as a strategic asset — to build authority, open doors, and create high-value opportunities — this conversation matters to you. Not because AI is coming for your book. But because the real issue underneath all the noise is one you should have been thinking about long before AI ever entered the picture: who actually owns your work, and what can be done with it.
What the Industry Is Saying
Walk the halls of any major publishing event right now and you’ll hear the same three notes on repeat.
Fear. AI is going to replace authors, editors, agents — pick your role, someone’s worried about it.
Copyright panic. Headlines about AI companies being sued, settled with, or investigated for how they trained their models have made “AI” and “stolen content” feel like synonyms in a lot of people’s minds.
“AI is stealing our work.” This is the emotional core of the conversation — a sense that something has been taken without permission, without payment, and without consent.
These reactions are understandable. They’re also, in large part, a reaction to a symptom rather than the underlying condition.
What They’re Missing
Here’s the reframe that almost nobody at these events seems willing to make: AI is a tool, not the threat.
A tool can be used well or badly. A tool can create enormous value or cause real harm, depending entirely on who’s holding it and under what terms. The printing press didn’t “threaten” authors — but it absolutely changed who held the power in publishing, and for centuries that power sat with printers and publishers, not writers. The internet didn’t “threaten” musicians — but it took until streaming royalty structures matured for artists to claw back any meaningful share of the value their work generated.
AI is simply the latest iteration of a very old story: a new technology arrives, and the question that actually matters is who controls the terms of its use.
The real issues aren’t “AI” in the abstract. They’re:
- Ownership — Who legally owns the work in question, and can that be proven?
- Control — Who decides where, how, and by whom that work can be used?
- Licensing — Under what terms, for what compensation, and for how long?
These three things have always mattered. AI hasn’t created them as issues. It’s made them impossible to ignore.
Your Take: This Is the Authority Moment
If you’re building a business around your expertise — and your book is a core part of that strategy — here’s what should actually be on your radar.
Retain your rights. Before you sign any publishing, ghostwriting, or platform agreement, understand exactly what you’re signing away. Many traditional and even self-publishing-adjacent contracts include broad rights grabs — covering not just the printed book, but audio, translation, derivative works, and increasingly, AI training and licensing rights. If you don’t know what’s in your contract on this point, find out now.
Control your distribution. Where does your content live? Who has the ability to copy, syndicate, repurpose, or license it without coming back to you? If the answer is “a platform I don’t control,” that’s worth sitting with for a moment.
Understand the value of your IP. Your book isn’t just a book. It’s a body of intellectual property that can be licensed, adapted, excerpted, translated, and — increasingly — used to train or feed AI systems, whether that’s your own custom GPT trained on your frameworks, or a third party’s model trained on a dataset that happens to include your work. Both of these are opportunities if you control them. Both are risks if you don’t.
This is the authority moment, because most people in the industry are still arguing about whether AI is good or bad. The strategic operators are asking a much better question: “What does my IP position actually look like, and how do I strengthen it?”
The Real Risk
Let’s be specific about what should actually concern you — because it isn’t “an AI might read my book.”
Signing away rights you don’t fully understand. This has been the single biggest risk to authors for decades, long before generative AI existed. Broad, perpetual, all-rights contracts; “work for hire” clauses; vague language around “future technologies and formats not yet invented” — these things have been quietly sitting in publishing contracts for years. AI has simply made the consequences of those clauses more visible and more immediate.
Platform dependency. If your authority, your audience, and your distribution all live inside someone else’s ecosystem — a publisher’s catalogue, a single platform’s algorithm, a third-party’s terms of service — you don’t actually control the asset you’ve built. You’re renting your own visibility. And the terms of that rental can change at any time, often in ways you only discover after the fact.
Neither of these risks is new. AI didn’t invent them. What AI has done is shine an extremely bright light on contracts and content practices that have been operating quietly in the background for a very long time — and that brings us to the example that’s been dominating headlines.
A Useful Case Study: The Anthropic Settlement
You’ve likely seen headlines about Anthropic — the AI company behind Claude — reaching a proposed $1.5 billion settlement, pending final court approval at the time of writing, with authors over how some of its AI training data was sourced. It’s worth understanding what actually happened here, because it’s a near-perfect illustration of the point above.
The court’s findings drew a clear distinction between two different things. Training an AI model on books that were legally acquired was found to be transformative and protected — not the issue. The issue was that a large portion of the training data had been downloaded from “shadow library” sites — places like Library Genesis — which had been hosting pirated copies of books for years, long before any AI company came looking.
In other words: those books were already stolen. The piracy infrastructure — the shadow libraries, the unauthorised scans, the unlicensed mirrors — existed well before generative AI was a mainstream concern. What AI training did was pull from that existing pool of pirated content at scale, and in doing so, made a problem that had been quietly tolerated by the industry for years suddenly impossible to ignore.
This is the pattern worth sitting with. The arrival of publicly accessible AI didn’t create a copyright crisis. It accelerated and amplified one that already existed — exposing how much unauthorised copying, unlicensed distribution, and unprotected IP had been sitting in plain sight for years, simply because no one had the tools or incentive to act on it at scale.
For authors and IP owners, the lesson isn’t “be afraid of AI.” The lesson is: the gaps in how your work is protected, licensed, and tracked were always there. AI just made them matter faster.
The Real Opportunity
Here’s where the conversation gets genuinely exciting — and where most of the fear-based commentary stops looking.
For authors and IP owners who do have a handle on their rights, AI represents one of the most significant opportunities the publishing world has seen in a generation.
A production accelerator. Drafting, editing, structuring, repurposing — the mechanical overhead of turning expertise into a finished, professional book has dropped dramatically. The barrier to producing high-quality work has never been lower.
A marketing engine. Repurposing a book’s content into articles, social posts, email sequences, lead magnets, and talks used to take a team. Now it takes a strategy and the right prompts — and a book that was already designed with this in mind (which, as covered earlier in this series, should be the case from the outset).
A research assistant. Validating ideas, synthesising research, stress-testing arguments, identifying gaps — AI tools can compress weeks of preparatory work into days, freeing you to focus on the parts of the process that genuinely require your expertise and judgement.
None of this is a threat to a well-positioned author. It’s leverage.
The Real Reframe
So let’s put this plainly.
This isn’t “AI vs authors.”
This is authors who understand leverage vs authors who don’t.
The authors and thought leaders who will benefit most from this next phase of publishing aren’t the ones shouting loudest about AI at industry events. They’re the ones quietly auditing their contracts, clarifying their rights, building IP strategies that account for licensing and distribution, and then using AI tools to move faster than everyone else who’s still stuck arguing about whether the technology should exist at all.
The technology isn’t going anywhere. The only question that matters is whether you’re positioned to benefit from it — or exposed to it.
Think Strategically About Your IP
If you’re using your book as a serious business asset — and not just a vanity project — now is the moment to get clear on three things: what rights you currently hold, what rights you’ve signed away (knowingly or otherwise), and what your strategy is for licensing, distribution, and protection going forward.
Start by asking three questions:
- What do I own?
- What have I licensed?
- Who else has permission to use, adapt, distribute, or monetise my work?
This isn’t about panic. It’s about positioning.
However, I do want to stress that this article is not intended to be legal advice, and copyright law varies by territory. If you are unsure what rights you hold, what you have licensed, or what your contracts allow, speak to an IP solicitor or rights specialist.
The authors who come out of this next phase ahead won’t be the ones who shouted about AI the loudest. They’ll be the ones who understood, early, exactly what they owned — and exactly what it was worth.
This is the fourth article in “The Future of Publishing” series.
#1: The Publishing Industry is Changing
#2: Direct-to-Reader Book Sales: How Smart Business Authors Are Leaving Money on the Table
#3: Your Book Doesn’t Exist If It Can’t Be Found: Book Discovery in the Age of AI
Sources / Further Reading
Reuters: US judge considers Anthropic’s $1.5 billion settlement of authors’ lawsuit. https://www.reuters.com/legal/government/us-judge-considers-anthropics-15-billion-settlement-authors-lawsuit-2026-05-14/
Associated Press: Anthropic wins ruling on AI training in copyright lawsuit but must face trial on pirated books. https://apnews.com/article/1e5cece51c2e4bd0bb21d94de2abb035
Official Settlement Website: Bartz, et al. v. Anthropic PBC. https://www.anthropiccopyrightsettlement.com/United States District Court Order: Case 3:24-cv-05417-WHA Document 231. https://copyrightalliance.org/wp-content/uploads/2025/06/Bartz-v.-Anthropic-Order.pdf
