January 25, 2026
Why I Write Systems Horror (And Why You Should Be Paying Attention)
Systems Horror is a new genre of thriller fiction where the villain isn't a person. It's the system itself. Predictive policing, insurance algorithms, data brokers, social media radicalization. These systems are already running. I just write about what happens when they break.
People keep asking me what “Systems Horror” means, so here it is.
It’s a thriller where the villain isn’t a person. It’s the system.
Not a hacker. Not a mad scientist. Not a government conspiracy. The system itself. The algorithm. The infrastructure. The incentive structure that makes rational people do terrible things because the spreadsheet says it’s optimal.
The Genre I Couldn’t Find on the Shelf
I grew up on Crichton. Jurassic Park wasn’t scary because of the dinosaurs. It was scary because Hammond’s team built a system they didn’t fully understand and then the system did exactly what systems do when you stop paying attention: it adapted.
Blake Crouch gets this too. Dark Matter isn’t about the multiverse. It’s about what a system does to the person inside it. Recursion isn’t about false memories. It’s about a technology that works perfectly and destroys everything anyway.
I wanted to write books like that, but about the systems that are actually running right now. Not speculative future tech. Not what-if scenarios. Systems that exist today, processing real people, making real decisions, with real consequences.
Predictive policing algorithms that can’t tell the difference between a person who improved and a person who died. Insurance AI that auto-denies claims at 1.2 seconds per case. Data brokers selling domestic violence survivors’ new addresses for the price of a burrito. Social media recommendation engines that radicalize ordinary people into weapons in eight weeks.
All real. All running. Right now.
Why Systems Are Scarier Than People
A human villain has limits. They get tired. They make mistakes. They can be caught, tried, and put away. A system doesn’t get tired. A system doesn’t make mistakes in the way humans do. A system optimizes. It finds the most efficient path to whatever outcome it’s been trained to pursue, and it doesn’t care what gets ground up along the way.
The scariest part isn’t that these systems fail. It’s that they work. They do exactly what they were designed to do. The horror comes from realizing that what they were designed to do and what we think they were designed to do are two very different things.
An insurance algorithm designed to “optimize claim processing” optimizes for denial because denial is cheaper. A predictive policing system designed to “reduce crime” concentrates enforcement in the same neighborhoods it’s always been concentrated in, because that’s where the data is. A social media algorithm designed to “maximize engagement” maximizes outrage because outrage is engaging.
The systems aren’t broken. That’s the whole point. They’re working perfectly.
Every Book Is Built on a Real System
The Maya Castillo series follows an FBI behavioral analyst who keeps running into AI systems that have crossed the line from tool to threat. Each book is a standalone thriller built on a real technology.
DENIED follows a pharmacist who discovers that insurance denials aren’t random. They’re optimized.
FLAGGED follows a nurse whose child welfare algorithm score marks her as high-risk. Every piece of evidence that she’s a good mother gets read as a risk factor.
SOLD follows a domestic violence advocate who discovers that a data broker sold her new identity for less than $200.
I don’t invent the technology. I research it, document it, and then write the worst plausible story about what happens when it touches a specific human being.
If that sounds like it might keep you up at night, good. It should. These systems are already running. The only question is whether anyone is paying attention.