The New AI Detection Scam Targeting Authors—and What to Do About It

Scammers have found a new way to frighten authors into handing over money. Here’s how to protect yourself.

Just when you thought the publishing scams couldn’t get any more inventive, along comes a new one.

Author and blogger Anne R. Allen recently reported receiving two emails from people claiming to represent an AI detection company.

According to the emails, the company had scanned all her books, blog posts, and social media content and determined that everything she’d written was generated by artificial intelligence.

The solution? Pay them a substantial sum to “clean up” her writing before she was exposed and cancelled.

You can read Anne’s account in her post, Report from Scamistan: The Latest Publishing Scams Targeting Writers.

There’s a particularly absurd detail in Anne’s story: approximately 99% of her published work predates the arrival of modern generative AI.  Apparently, the scammers hadn’t considered that little complication.

But while Anne could recognize the scam for what it was, she also understood why it might frighten other writers.  And that’s what makes this particular scam so nasty.

The Scammers Have Found a New Fear to Exploit

Publishing scams aren’t new. Authors have been approached by fake literary agents, bogus publishers, imaginary film producers, and dubious marketing companies for decades. The underlying strategy rarely changes:  Convince the author that something wonderful—or terrible—is about to happen, then demand money to make it happen or prevent it.

The new AI detection scam follows that familiar pattern, but it exploits a particularly sensitive issue in today’s publishing industry.  Authors are increasingly worried about being accused of using AI to write their books.

Some publishers and agents are using AI detection software to screen manuscripts. Authors have reported false accusations. Debates about whether AI detectors are sufficiently reliable to make professional decisions have become increasingly heated.

And scammers have spotted an opportunity. They don’t need to prove that you’ve used AI. They don’t even need to run your writing through a detector.  They only need to convince you that someone else might believe their accusation.

This isn’t really an AI detection scam. It’s a reputation-extortion scam using AI detection as the threat. And it could be particularly frightening for authors whose livelihoods depend on maintaining a professional reputation.

There Are Two Different Problems Here

Before deciding how to protect yourself, it’s important to distinguish between two very different situations.

The first is an unsolicited accusation from a stranger who wants money.

The second is an accusation from someone who has a legitimate professional relationship with you: a publisher, literary agent, retailer, contest administrator, or freelance client.

Both may involve AI detection software, but the appropriate responses are completely different.

Situation One: A Stranger Claims Your Books Are AI-Generated

Suppose an email arrives claiming that a company has analyzed your books and discovered AI-generated content. The sender warns that your reputation is in danger, that your books could be removed from sale, or that you’ll be publicly exposed.

Fortunately, they can fix everything….for a fee.

Here’s what to do.

1. Don’t reply.

Don’t explain that you wrote the book yourself. Don’t tell them you published it in 2009. Don’t argue about their findings or demand to see their evidence. The sender isn’t interested in establishing the truth. They’re interested in establishing whether you’re frightened enough to pay.

Even an indignant response tells them they’ve reached a real person who takes the accusation seriously.

2. Don’t click links or open attachments.

The supposed detection report could be a phishing link, a malicious attachment, or a route to a fraudulent payment site. Don’t upload your manuscript to a service the sender recommends, either. There’s no reason to hand your unpublished work to a stranger who has just threatened you.

3. Don’t pay for a cleanup service.

This is the heart of the scam. The sender creates a problem, then offers to sell you the solution. There is no reason to believe the alleged problem exists, and no reason to believe the proposed service would solve it.

4. Preserve the evidence, then block and report.

Keep a copy of the threatening email, preferably with its full headers, in case you need to report the attempt. Mark it as phishing or spam and block the sender.

If the threat involves actual extortion, impersonation, or threats to contact your publisher or readers, consider reporting it to the appropriate fraud authorities. In Canada, that includes the Canadian Anti-Fraud Centre.

5. Verify any claims independently.

If the email claims to come from your publisher, a retailer, or an established detection company, don’t use the contact information supplied in the message. Go directly to the organization’s official website or your existing business contact and ask them directly about the communication.  An unfamiliar sender claiming to represent a familiar company is still an unfamiliar sender, and scammers pretending to be publishing professionals are everywhere.

And remember: a demand for money to prevent exposure is a major warning sign, regardless of whose name appears at the bottom of the email.

Situation Two: A Publisher or Agent Questions Your Manuscript

This situation is considerably more complicated. Suppose your publisher contacts you to say that an AI detector has flagged portions of your manuscript. Or an agent reviewing your submission says the manuscript appears to contain AI-generated material.

This isn’t necessarily a scam.

It may be a genuine professional concern, even if the detection result is wrong. Ignoring the accusation could damage a legitimate business relationship. But neither should you panic.

1. Preserve the manuscript exactly as it was submitted.

Don’t immediately start rewriting passages to make them appear more human. Keep the original file, with its dates and metadata, along with any previous drafts. You need an accurate record of what was submitted and when.

2. Ask for the specifics.

Which passages were flagged?  Which detection software was used?  What did the report actually conclude?  What policy or contractual requirement is supposedly involved?

These are reasonable questions.  An AI detector’s assessment is not the same thing as proof of authorship.

3. Review the applicable policy.

Publishers have different rules about AI use. Some prohibit AI-generated manuscript text. Others distinguish between generated prose, research assistance, editing, brainstorming, and administrative uses.

Don’t assume that one publisher’s policy applies everywhere. Find out precisely what your publisher prohibits, and whether that prohibition was part of your agreement.

4. Offer relevant evidence of your writing process.

If necessary, you may be able to provide earlier drafts, outlines, research notes, version histories, editorial correspondence, or other records showing how the manuscript developed. 

You don’t necessarily need every scrap of paper you’ve ever scribbled on.  A reasonable collection of ordinary working files may be enough to resolve a misunderstanding.

5. Escalate if necessary.

If the accusation threatens your contract, publication, payment, or professional reputation, involve your literary agent, authors’ organization, or a qualified publishing lawyer. 

Don’t agree to contractual changes, admissions, or costly remedial work simply because a detection report has been presented as authoritative.

The stakes may be too high to handle the dispute alone.

The Authors Guild Has Something to Say About This

In October 2026, the Authors Guild published AI Best Practices for Publishers: Using AI Detectors and Dealing With Undisclosed AI Use.  The guidance makes several important recommendations.

Publishers and agents should not rely solely on AI detectors to determine whether a manuscript contains AI-generated text. 

Authors should be consulted when a detector raises concerns.

Publishers should have clear, transparent AI policies.

And, perhaps most importantly, the Guild argues that authors should not be required to prove that they wrote their own books.

That’s an important principle.  The burden shouldn’t automatically fall on authors to establish their innocence every time a software program produces a suspicious score.

The Guild also opposes mandatory technologies that track or record the writing process simply to prove human authorship.

However, it supports authors who voluntarily preserve records of their work in case a dispute arises.

Which brings us to something every author can do right now.

Protect Yourself Before Anyone Makes an Accusation

You don’t need elaborate technology, special software, or a complicated documentation system. You need good working habits.  And most of these are habits productive authors should already have.

Keep your drafts.

Don’t overwrite your only copy of a manuscript every time you revise it. Save major versions separately, or use software that maintains version histories.

Keep your planning material.

Outlines, character notes, timelines, worldbuilding documents, and research files all help establish the development of a project.

They’re also useful when writing sequels, revising older work, or returning to an abandoned manuscript.

Back up your work.

Maintain regular backups, preferably including a copy stored separately from your primary computer.

A backup system that preserves older versions is especially useful.

Keep relevant correspondence.

Editorial emails, beta-reader feedback, and discussions with your publisher can help establish the history of a manuscript.

You don’t need to archive every casual conversation, but don’t casually discard important project records.

Keep your contracts and AI policies.

If you’re working with publishers, agents, or freelance clients, retain copies of the rules and agreements that applied when the work was submitted.  Policies can change. You need to know what you actually agreed to.

None of this requires changing how you write. It simply means preserving the ordinary records your writing process already produces.

And those records have value well beyond defending against an AI accusation.  They protect your intellectual property, support your business, and make it easier to manage a long writing career.

Don’t Rewrite Your Books to Please an AI Detector

There’s another trap authors should avoid. The growing market for services promising to make writing appear more human.  Some offer to rewrite text so it will pass AI detection software. Others promise to reduce an AI score or remove supposedly suspicious language.

But consider what that means for an author who actually wrote the manuscript.

• You wrote the book.
• A machine claims a machine wrote the book.
• So you change your own writing until another machine decides it looks sufficiently human.

What exactly have we accomplished?  Even if the revised text passes one detector, there’s no guarantee it will pass another.

And the revisions may have weakened the author’s voice, damaged the prose, or altered perfectly good writing for no legitimate reason.

Don’t let a detector become your editor.

If a legitimate publisher raises concerns, address those concerns through the normal professional process.  But don’t start feeding your manuscripts into every available detector and rewriting sentences until the scores look reassuring.  That’s not a productive use of your writing time.

Don’t Let Fear Change the Way You Write

This may be the most important recommendation of all.

Authors already have enough to worry about. We’re managing production schedules, marketing, publishing platforms, retailers, contracts, technology, and the endless business of getting books into readers’ hands.  Now we’re supposed to worry that software might decide we didn’t write our own books?

And that strangers might use those decisions to threaten our reputations?

No.

We should be aware of the risks, certainly.  We should understand our contractual obligations, keep sensible records, and know how to respond if an accusation arises. But we shouldn’t reorganize our creative processes around the possibility of being falsely accused.

Nor should we give scammers the power to make us question our own work.

The solution isn’t to become obsessed with proving our humanity. It’s to recognize the difference between a fraudulent threat and a genuine professional issue, and respond appropriately to each.

Your AI Accusation Action Plan

Here’s the short version to keep handy.

If a stranger accuses you of using AI and demands money:

• Don’t respond, click links, or send manuscripts.
• Don’t pay for detection, cleanup, or humanization services.
• Preserve the message.
• Block and report the sender.
• Independently verify any claim involving a real publisher or retailer.

If a legitimate business partner questions your work:

• Preserve the exact manuscript that was questioned.
• Ask for the flagged passages, detector results, and relevant policy.
• Review your contractual obligations.
• Gather existing drafts and writing records if needed.
• Seek professional advice if your contract or reputation is at risk.

For everyday protection:

• Keep drafts, notes, and version histories.
• Maintain reliable backups.
• Retain important publishing correspondence and agreements.
• Don’t rewrite authentic prose merely to satisfy a detector.
• Continue writing in the way that works best for you.

Anne R. Allen’s experience is a timely warning that scammers are paying attention to the controversies affecting our industry.  They’re adapting their tactics to exploit whatever frightens authors most.

We can’t stop them from sending emails.  But we can make sure those emails don’t work.  The best defense against this particular scam is knowing that you don’t have to buy your way out of an accusation someone invented.

And the best preparation for a genuine dispute is something you should be doing anyway.

Keep good records.

Back up your work.

Then get on with writing the next book.

---

Have you received an email claiming your books were AI-generated? Or have you encountered a genuine AI detection dispute with a publisher, retailer, or other industry professional? I’d be interested to hear about your experience in the comments.

References and Further Reading

• Anne R. Allen, Report from Scamistan: The Latest Publishing Scams Targeting Writers, October 31, 2026.
• The Authors Guild, AI Best Practices for Publishers: Using AI Detectors and Dealing With Undisclosed AI Use, October 13, 2026.
• The Authors Guild, AI Best Practices for Authors, May 11, 2026.
• Canadian Anti-Fraud Centre.

Write More, Faster Than Ever Before | Are You Prolific?

Editing Your Next Novel? Mark Posey offers fast, writer-friendly edits with zero drama. Check out services »