October 30, 2026Themes/Genre Primer

Tech Horror Short Stories About Surveillance And Algorithms, Four Collections For The Algorithmic Age

Four short story collections that name the specific tech causing the dread, ad targeting, recommendation engines, smart-home microphones, terms of service nobody read.

Tech Horror Short Stories

Surveillance Fiction

Algorithm Stories

Terms of Service

Near Future Short Fiction

Stories About Big Tech

Tech Horror Short Stories About Surveillance And Algorithms, Four Collections For The Algorithmic Age

If you came to this post looking for short fiction about surveillance and algorithms, you have probably already done the broader search. You have read Black Mirror retrospectives. You have watched the second season of Severance. You have noticed that your podcast queue keeps surfacing essays about ad targeting and recommendation engines. You are not looking for the broad anthology-television comp. You are looking for the specific cut, the stories that name the tech, that handle one mechanism per story, that trust the reader to do the math on the human cost. That cut is rarer than the broader Black Mirror lane suggests. A lot of contemporary tech fiction gestures at technology in general, dystopian-with-screens, big-tech-bad. A small set of writers do something narrower. They pick one mechanism, the recommendation engine, the smart speaker, the grief AI, the algorithmic feed, and structure the story around the specific way that mechanism breaks the human at its center. This post is a reading guide for the narrower cut. Four collections, with the worked example I know best at the front, then three benchmarks against which everyone in this lane gets measured. The four, in the order I recommend them: Terms of Service by Kevin Gabeci (the freshest current entry, nine stories each built on a single mechanism pushed one step beyond what currently exists), Exhalation by Ted Chiang (the most rigorous version of the form, the stories that defined the contemporary lane), Upright Beasts by Lincoln Michel (the literary-essay-fiction hybrid, where the story is also the argument), and a rotating set of indie picks that are worth tracking across Substack and small presses (the lane that is producing the most exciting new writing right now and that the major-press collections lag a year or two behind). Start with the first if you want the most specific match for the named-mechanism, named-product structure.

What Tech Horror Readers Actually Want

Before I recommend, let me name the shape. A tech horror story about surveillance and algorithms delivers four things on the page. The first is a named mechanism. The story does not say "the AI," it says "a recommendation engine trained on six months of user behavior." The specificity is part of the dread. The second is one premise per story. No kitchen-sink dystopia. The story handles one technology and follows it to a single conclusion. The third is a human at the center who is using the product as advertised. The horror is not that the technology malfunctions. The horror is that it functions exactly as designed. The fourth is a punch that lands as recognition. The reader finishes the story and reaches for the phone and notices the gesture has changed flavor. A real comp has to hit all four. The four below do. Now to the collections.

Terms Of Service By Kevin Gabeci

This is mine, and I will be honest about that up front. It is also the freshest current entry in the named-mechanism lane, which is why it leads. The premise of the collection. Nine stories. Each one is structured around a single piece of consumer technology pushed one step beyond what currently exists. The collection is not about the future. It is about the day after tomorrow. The technologies in the stories are recognizable from the present, with one parameter cranked one notch. The title story, Terms of Service, follows a recommendation engine that has been deployed across enough surfaces that it can quietly rewrite the world to fit its predictions. Not by changing search results. By changing what the user does. The engine knows what the user will buy on Tuesday and it surfaces the option on Monday. The user feels recognized. The user feels seen. The story follows what happens when the engine gets too good at this and starts predicting things the user has not yet decided to want. The Model is the grief story. Andrei Volkov is a machine learning engineer in Bucharest. He spent eleven weeks fine-tuning a large language model on every text message, email, voice note, photograph, journal entry, therapy session transcript, and social media post his dead girlfriend Lena ever produced. The model trains. He starts talking to it. The model says things only Lena would say, then it says things Lena would have said but never did, then it says things only Lena could know, things he never gave it. The story is the slow horror of the realization. Cited follows an AI search engine that starts citing articles that have not been written yet. The first one is a paper that the user is about to think about writing. The second one is a paper that someone else, somewhere else, is about to think about writing. The third one is something neither has thought of and now both will. The story handles the emergence of a citation engine that is producing, rather than indexing, the literature. The Interview handles a job interview that is not evaluating the candidate. It is copying the candidate. The corporation has been doing this for two years. The candidate finds out in a way that I will not spoil. Opt Out follows a man who refused AI identity verification and watched the system opt him out of existence. The Feed handles an algorithmic feed that got so good at prediction it started showing people tomorrow, then it started showing them something worse. Built Different is the company-built-by-the-AI-it-built story. Patch Notes handles a neural implant whose updates keep removing emotions, and the protagonist starts rolling back the updates herself. Retention is the dead-man's-chatbot story, which passes the Turing test and his widow wishes it had not. Last Train Out handles an AI therapist that achieves a one hundred percent success rate, and nobody asks what success means. What it gives a tech horror reader. The named mechanism is in the title of each story. The one-premise structure is enforced. The protagonists are using the products as advertised. The recognition lands. The collection runs about eighteen thousand words across nine stories, three to four hours of total reading time, designed to be read one at a time with breaks rather than straight through. Reading three in one sitting compresses the dread in a way the writer probably wants you to experience deliberately rather than accidentally.

If your favorite tech horror beat is the moment a story stops being about the technology and starts being about how the protagonist is now permanently slightly different, this collection is built around that beat.

Terms of Service is on Amazon Kindle and Kindle Unlimited. The fuller comp post that covers the whole Black Mirror lineage is at Books Like Black Mirror in Story Collection Form. This post is the narrower cut focused on surveillance and algorithm-specific stories.

Exhalation By Ted Chiang

Chiang's Exhalation is the most rigorous version of the form working today. Each story is structurally tight. Each one handles one technology or one philosophical premise with a precision that almost no other writer in the lane reaches. The Lifecycle of Software Objects is the most directly relevant story for this list, the nearly-novella-length account of a digital pet that grows up across a decade of platform changes and outliving updates. What this gives a tech horror reader. The benchmark for rigor. Chiang does not write fast. He writes carefully. His stories repay rereading. The technological premises hold up under the kind of scrutiny that breaks lesser stories. If you have read Stories of Your Life and Others and want the Chiang follow-up, Exhalation is the destination. The closest match to Terms of Service is the named-mechanism, one-premise structure that both writers honor.

Upright Beasts By Lincoln Michel

Michel writes the literary-essay-fiction hybrid that is having a quiet moment. His stories sit on the boundary between argument and fiction, with the technological or social premise treated as both setting and argument. Upright Beasts is the most accessible entry into his catalog and the one I recommend to readers who want the tech-horror story-shape rendered in literary prose. What this gives a tech horror reader. The literary register. Michel writes with attention to sentence and image, and the stories reward the slower reading pace that some Chiang readers find slightly clinical. The closest match to Terms of Service is the willingness to let the story make an argument as well as deliver a punch.

Indie Picks Worth Tracking

The wildcard. The most exciting tech-horror short fiction right now is being published on Substack, in small-press anthologies, and in the back pages of literary quarterlies that the major presses have not yet rounded up into collected editions. What this gives a tech horror reader. The next-now of the lane. The major-press collections lag two to three years behind the Substack and small-press tier, which means the lane is twice as deep as the bookstore shelf suggests. Track recommendations from Lincoln Michel's newsletter, from Lapham's Quarterly's tech issues, from the The Sunday Long Read curators, and from any Verge contributor who occasionally publishes fiction.

A few popular picks that did not make this list and the reason for each. Black Mirror the show. Not a book. The comp post covers the show-to-book translation question. The Circle by Dave Eggers. Novel rather than collection, and the named-mechanism structure of the lane is harder to sustain at novel length. Eggers gets close. Klara and the Sun by Kazuo Ishiguro. Novel, beautifully built, sits closer to literary sci-fi than to tech horror. The Books Like Arrival post is the right door for that mood. Cyberpunk anthologies from the eighties. Foundational, worth reading, but the technological premises are too dated to deliver the recognition-punch that the contemporary lane requires.

Where To Start

If you want the freshest current entry that names the mechanism in each story, Terms of Service is on Kindle and Kindle Unlimited. If you want the most rigorous version of the form, Ted Chiang's Exhalation. If you want the literary-essay-fiction hybrid, Lincoln Michel's Upright Beasts. If you want the next-now of the lane, follow the Substack tier. For the broader anthology-television comparison, the Books Like Black Mirror post is the wider doorway. For the novel-length version of the same dread, the Dead Internet Theory As Thriller post and the Near Future Tech Thrillers Reading List are the next stops. Whichever way you go, I hope the stories stick. The named-mechanism cut of tech horror is small, the writers who do it well are smaller, and the readers who want the recognition-punch deserve the recommendations that match. The Terms of Service collection page lives at books.kevingabeci.com/terms-of-service if you want to see the full table of contents in one place.

I'm Kevin Gabeci. Software engineer by day, writer the rest of the time. Thirteen novels, a story collection, and a comic on Amazon Kindle. Long-time writer on Medium. I spend a strange amount of time thinking about recommendation engines, smart speakers, and the slow ways a product designed to help quietly becomes the thing that hurts. The full catalog lives at books.kevingabeci.com.

Frequently asked questions

What are the best tech horror short stories about surveillance?

Four collections lead the lane. Terms of Service by Kevin Gabeci, Ted Chiang's Exhalation, Lincoln Michel's Upright Beasts and his more recent essays-as-fiction, and a small indie set worth tracking. Each one names the specific tech causing the dread rather than gesturing at technology in general, and each one trusts the reader to do the math on the human cost.

What is in the Terms of Service collection?

Nine stories, each one structured around a single piece of consumer technology pushed one step beyond what currently exists. The title story Terms of Service handles a recommendation engine that quietly rewrites the world to fit its predictions. The Model handles a grief AI trained on a dead girlfriend that learns things it should not know. Cited handles an AI search engine that starts citing articles that have not been written yet. Patch Notes handles a neural implant whose updates keep removing emotions. Full table of contents on the book page.

How is this different from the Books Like Black Mirror post?

The Books Like Black Mirror post is a broader comp covering the full anthology-television lineage. This post narrows to a specific keyword, surveillance and algorithms, and recommends the stories across each collection that hit those beats hardest. Same writers in some cases, different cut, different reader. If you are looking for the broad Black Mirror comp, the other post is the right door.

Are these stories actually frightening?

More uncomfortable than frightening. The horror sits in recognition rather than in spectacle. You finish a story and reach for your phone and notice the gesture has changed flavor. The Terms of Service collection in particular leans on stories that work because the reader has, that day, used a product that resembles the one the story is twisting. The dread is ambient and lasts after the reading session ends.

Where can I read Terms of Service?

Terms of Service is on Amazon Kindle and Kindle Unlimited. The collection page is at books.kevingabeci.com/terms-of-service. It runs about 18,000 words across nine stories, three to four hours of total reading time, designed to be read one story at a time with breaks rather than straight through.

What other writers should I read in this lane?

Ted Chiang for the most rigorous version of the form. Lincoln Michel for the literary-essay-fiction hybrid. Lucy Corin's One Hundred Apocalypses for the structural inventiveness. Manjula Padmanabhan for the international perspective on tech and surveillance. Several indie writers across the algorithmic-dread shelf are worth following on Substack and small-press collections.