Everybody Wants a Cut, and Some of Them Aren’t Even Real

When money finally shows up in a fight authors have been having with AI companies, who do you think arrives first — the writers, or the people holding a piece of paper that says they’re owed something?

That’s the uncomfortable question sitting underneath the pushback happening right now. Authors are resisting claims from publishers and agents on the Anthropic settlement, and at the same time, scams aimed at writers are climbing. Those two things are not unrelated. Any time a payout becomes public knowledge, two categories of people start sending emails: the ones with a legitimate contractual argument, and the ones who just found out your name is attached to a check.

I review tools for a living. I spend my days figuring out whether a piece of software does what its marketing page claims. So my angle here isn’t legal — it’s practical. Verification is a tooling problem, and most writers are running with almost no tooling at all.

The scam surface just got wider

Writer Beware has documented how direct solicitation became one of the main ways scammers recruit victims, going back to the first overseas waves in the mid-twentyteens. The other main channel is ads. Neither of those requires anything sophisticated. It requires a list of names and a plausible reason to contact you.

A settlement gives them the most plausible reason they’ve had in years. And the impersonation attempts have already gotten bold — authors have been receiving fraudulent emails from someone posing as the United States Copyright Office, asking recipients to verify their copyright registration. Think about how well that pretext fits the current moment. You’re an author who may have a claim tied to your registered work. An email arrives from what looks like the office that holds that registration. It asks you to confirm details. The framing does all the work.

The University of North Carolina Press has been posting tips because scams targeting authors keep rising. When university presses start running public safety advisories for their own writers, the volume is real.

What actually works for verification

Here’s my honest read after years of testing workflows: the tools that help are boring, and the tools that promise to do your thinking for you are the ones to avoid.

• Out-of-band confirmation. If an email claims to be from your publisher, your agent, or a government office, close the email. Go find the phone number or contact page independently. Never reply, never click. This single habit defeats most impersonation attempts, and it requires no software.
• A real database instead of your inbox. Writer’s Digest named Duotrope one of its best agent websites for writers for 2026, and paid subscribers get a searchable database of current fiction, poetry, nonfiction, and visual art publishers and agents, plus a deadline calendar and a submission tracker. The value here isn’t discovery. It’s having an independent record of who exists and who represents what, so an unsolicited claim can be checked against something other than the claimant’s own signature block.
• Your own contract file. Every claim from a publisher or agent traces back to language in an agreement you signed. If you can’t find those agreements in under five minutes, that’s the first tool gap to close. A folder and consistent file names beat any AI assistant here.
• Slow down on anything urgent. Urgency is the tell. Legitimate contractual disputes move through lawyers over months. Scams need you to act today.

Where AI tools help, and where they don’t

I’ll be blunt about the category I cover. An AI summarizer can help you understand dense contract language faster. It can flag clauses about subsidiary rights or electronic reproduction that you’d otherwise skim past. That’s a legitimate use, and it’s a good one.

What it cannot do is tell you whether the person emailing you is who they say they are. No language model verifies identity. It has no access to the registry, no ability to call your agent, no way to confirm a domain belongs to who it claims. If a tool markets itself as protecting writers from publishing scams, ask what data source it checks against. If the answer is vague, the product is vibes.

There’s also a quieter risk. Pasting your unpublished manuscript or your signed contracts into a free consumer chatbot to “check if this is a scam” hands your material to a third party whose retention policy you probably haven’t read. That’s an odd move in a moment defined by an argument over who gets to use authors’ text.

The part that matters

Authors pushing back on claims from publishers and agents is a fight over contract interpretation, and it deserves to be argued properly. But the fight has created noise, and noise is cover. Every writer following this story is now a warmer target than they were six months ago.

Treat every inbound claim on your work as unverified until you’ve confirmed it through a channel you chose. Not the one that chose you.

🕒 Published: September 7, 2026