The Publishing Industry Has an AI Problem

Literary agents are drowning. Since late 2024, submission queues at major agencies have swelled by 50 to 300 percent depending on who you ask. Not because more people suddenly decided to write novels. Because AI made it trivially easy to generate a manuscript-length document and click "submit."

Some agencies now include a checkbox in their submission forms: "I confirm this manuscript was written without the use of generative AI." A few have gone further, requiring signed declarations. One well-known agency added a clause to their submission guidelines stating that any manuscript found to be AI-generated will result in a permanent ban from future submissions.

The intention is good. The enforcement is nearly impossible.

Declarations Don't Prove Anything

A checkbox is just a checkbox. A signed declaration is just a signature. Neither provides any actual evidence that a manuscript was written by a human being. They are trust-based mechanisms in an environment where trust is exactly what's being eroded.

AI detection tools are unreliable. They produce false positives on human-written text and can be fooled by light paraphrasing. No major publisher has adopted them as a gating mechanism for submissions, and for good reason. The liability of rejecting a genuinely human-written manuscript based on a flawed algorithm is significant.

So what can an author who actually wrote their book do to prove it?

The Evidence Problem

Think about what a manuscript typically looks like when it arrives at an agent's desk: a Word document. Maybe a PDF. A clean, finished product with no history, no context, no trail showing how it came to exist.

Whether a human spent two years writing it or an AI generated it in forty-five minutes, the final artifact looks the same. A Word file is a Word file.

This is the core problem. Proving human authorship requires evidence that exists before the final document: evidence of the process, not just the product.

Two Paths to Provable Authorship

There are two practical ways for an author to build a verifiable chain of evidence that their manuscript is human-written.

Path 1: Write by Hand or Typewriter, Then Digitalize

Many literary writers already compose first drafts by hand. Cormac McCarthy wrote on a typewriter. Neil Gaiman writes first drafts in notebooks. This is not unusual.

Physical pages are powerful evidence. Handwriting carries biometric characteristics (letter formation, spacing patterns, pressure variations) that are unique to the individual and essentially impossible to fabricate at manuscript scale. A typewriter leaves its own forensic fingerprint: specific character misalignments, ink density variations, mechanical quirks unique to that machine.

When you scan these pages with LyteWriter, the AI extracts the text while the platform records the relationship between the physical source and the digital output. Each page gets a Seal of Humanity, a cryptographic certification that the content originated from a verified human source. The scanned images, the extracted text, and the Seal together form an evidence chain from notebook to finished manuscript.

For a 300-page novel written across several notebooks, that is 300 individually sealed pages, each traceable to a physical source. That is not a checkbox. That is evidence.

Path 2: Write in LyteWriter's Editor with Keystroke Dynamics

Not every author writes by hand. Some type their first drafts, but they type them themselves, with their own fingers on a keyboard.

When you write in LyteWriter's editor, the platform analyzes your keystroke dynamics: typing rhythm, pause patterns, error-and-correction behavior, the cadence of composition versus editing. These behavioral biometrics are as unique as a fingerprint and build up over time into a profile that is extremely difficult to fake.

Each chapter or section written in the editor earns its own Seal of Humanity based on this behavioral analysis. Over the course of a manuscript, you accumulate a portfolio of Seals, one for each writing session or document, that collectively demonstrate sustained human authorship across the entire work.

AI-generated text pasted into any editor has no keystroke dynamics. It appears instantaneously. The behavioral signature is absent. That difference is measurable and verifiable.

Building a Portfolio of Proof

The practical workflow for an author concerned about proving authorship looks like this:

For handwriters: Write in your notebooks as you always have. Periodically scan your pages with LyteWriter. The platform extracts the text, organizes it in nested folders (by chapter, by draft, however you work), and seals each page. When the manuscript is done, export the final version as a DOCX or PDF and submit it along with a note that every chapter carries a verifiable Seal of Humanity, checkable at lytewriter.com/verify by anyone, no account needed.

For typists: Write your manuscript in LyteWriter's editor. The keystroke analysis runs in the background; you do not need to do anything differently. Each writing session builds your behavioral profile and contributes to the Seal for that document. When you are ready to submit, export and include the same verification note.

For hybrid writers (handwritten first draft, typed revision): You get both layers of evidence. The scanned first draft shows the original handwritten composition. The typed revision in the editor shows the human editing process. Two independent streams of proof for a single manuscript.

What Agents and Publishers Can Do with This

The verification is public. An agent who receives a manuscript with Seal of Humanity references can visit lytewriter.com/verify, enter the verification code, and see the certification status immediately. No software to install. No account to create. No fee to pay.

This does not require the agent to trust the author's word. It does not require the agent to trust an AI detection tool. It requires the agent to check a cryptographic record that is either valid or it is not.

Some publishers are beginning to explore provenance requirements for submitted manuscripts. The infrastructure for verifiable human authorship exists now. The question is how quickly the industry will adopt it.

This Is Not Surveillance

To be clear about what this is and what it is not.

This is not a system that monitors writers or restricts how they work. LyteWriter does not watch you write in other applications. It does not scan your hard drive. It does not report your activity to anyone.

This is an opt-in tool for writers who want to build proof of their own authorship. You choose to scan your pages or write in the editor. You choose to include verification codes in your submissions. You control the evidence.

The writers who will benefit most are the ones who need it least: honest authors who actually wrote their manuscripts and simply want a way to demonstrate that fact in an environment where their word alone is no longer sufficient.

In a publishing world increasingly skeptical of what is human and what is not, that proof is becoming valuable. Not because anyone should have to prove they are human. But because those who can will stand apart from those who cannot.

See also how journalists use the Seal to verify field notes, and for fiction writers specifically, our guide to writing your first draft by hand.

Get started with LyteWriter for free. 10 scans per month, no credit card required.