Your Boss Bought a Robot That Guesses the Lowest Salary You’ll Say Yes To
Your Boss Bought a Robot That Guesses the Lowest Salary You’ll Say Yes ToTurns out your payday loans, your Google searches, and your maxed-out credit card are being sold to figure out exactly how little they can pay you. Cool cool cool.
Signals being sold: payday-loan history, credit-card balances, location data, your search habits, even social media vibes. One state (Colorado) is trying to straight-up ban it.
Full breakdown from Cory Doctorow’s Pluralistic and the UC Irvine Law writeup of the Washington Center for Equitable Growth report.
OKAY SO. You know how websites sometimes charge YOU a different price than your friend for the same flight? (It’s called surveillance pricing, and it’s gross.) Well someone in an office looked at that and went “what if we did that… but for people’s paychecks.” And now it’s a real thing. It’s called surveillance wages and honestly reading about it made me want to flip a table.
Here’s the whole scam in one sentence: companies quietly buy data about how broke and desperate you probably are, feed it into a program, and that program spits out the smallest number you’ll likely accept. Then they offer you a tiny bit above that. You feel like you “won” the negotiation. You did not win the negotiation (I’m sorry).
🧩 Dumb Mode Dictionary — every weird word, in plain English
| Fancy term | What it actually means |
|---|---|
| Surveillance wages | Using spy data about you to pay you as little as possible |
| Data broker | A company that secretly collects your info and sells it. You never signed up. |
| Algorithm | A recipe a computer follows. Input = your data. Output = your lowball number. |
| Financial vulnerability signal | Proof you’re stressed for money (payday loan, high card balance) = they smell blood |
| ATS | “Applicant Tracking System” — the robot software that reads your job application first |
| Wage bill | The total money a company spends on paychecks. Shrinking it = their bonus. |
😤 So how does this even work?
Simple and evil:
- A program pulls whatever it can find on you — did you take a payday loan? Is your credit card near the limit? Where do you live? What have you been Googling?
- All of that gets collected without you knowing, usually bought from data brokers.
- The more “desperate for cash” you look, the LOWER the number it tells your future boss to offer.
- Offer you a dollar above your secret floor, and you’ll probably take it — instant savings on their wage bill.
It’s the same trick as the app that charges you more for a ride when your phone battery is low. Same energy. Different victim (you, at work).
📊 The receipts — what they're actually looking at
| Signal they buy | What they think it says |
|---|---|
 Payday loan / high card balance |
“This person needs money NOW, lowball them” |
 Location data |
“Lives in a cheap area, offer less” |
 Google search history |
“Been searching ‘rent help’ — they’re desperate” |
 Social media activity |
“Job-hunting for weeks, running out of options” |
None of this measures whether you’re good at the job. Zero. It only measures how much you can be squeezed. (Doctorow calls the whole thing “empiricism-washing” — dressing up a shakedown as fancy math.)
🗣️ What the timeline's saying
- Over on Hacker News people are pointing out the wildest part: it’s basically legal right now in most places.
- Colorado actually wrote a bill — the “Prohibit Surveillance Data to Set Prices and Wages Act” — to ban companies from using your payday-loan history, location, or search behavior to set pay. (One state. So far.)
- Labor researchers have been screaming about “algorithmic wage discrimination” for years — gig drivers already live this, where two people doing the identical delivery get paid different amounts based on how likely each is to quit.
🧠 Why a senior engineer would go 'huh'
The genius-evil part: the system doesn’t need to be accurate. It just needs to be accurate on average. If it lowballs 100 people and 60 accept, the company still saves money overall — even if 40 walk. Your individual dignity is a rounding error in their spreadsheet. That’s the quiet horror of doing anything “at scale.” The fix isn’t better data — it’s making the desperation signal impossible to buy in the first place. Which, conveniently, is where the money is now. 
Cool. So a Spreadsheet Thinks You’re Cheap… Now What the Hell Do We Do? (╯°□°)╯︵ ┻━┻
Here’s the thing nobody’s built yet: if companies pay to READ these signals, there’s a whole business in helping people CONTROL what those signals say. The data flows one way right now. We reverse it. Five plays before the suits patch this:
🧹 The Signal Scrubber
Companies buy your “you’re broke” trail from data brokers. So… erase the trail before you apply. There are free opt-out forms at brokers like Acxiom, LexisNexis, and dozens more — but they’re a boring maze nobody finishes. You do it FOR people, right before their big interviews. Bundle it as a “pre-interview cleanup.”
Example: A 24-year-old virtual assistant in the Philippines posts in Facebook job-seeker groups, sells a $9 checklist + a “done-for-you” opt-out service for $40. Walks each client through broker opt-outs and resets their location/search signals the week before interviews. 60 clients a month.
Timeline: First paying client in ~1 week (people are hungry for this). Realistically the brokers make opt-outs harder within 6–9 months, so ride it while the forms are still easy.
🎭 The Fake-Rich Ghost
You can’t just delete signals — you can also add good ones. Poison the well. If the algorithm thinks you’d walk away from a lowball, it tells them to offer more. Keep your credit-card use low the month before applying, follow “I have options” type stuff, look un-desperate on paper. Sell the playbook of WHICH signals matter most.
Example: A 27-year-old in Lagos sells a one-page “signal glow-up” PDF on Gumroad for $12 — a plain-English list of what brokers weigh heaviest and how to look less squeezable. 300 sold = $3,600.
📓 The Snitch List
Nobody has a public list of WHICH employers and hiring-software vendors use these creepy wage algorithms. Be the dictionary. Crowdsource it — free to read, paid to get the full searchable version + alerts. First good list becomes THE page everyone links to.
Example: A 22-year-old in India runs a free Airtable database where job seekers report suspected “surveillance wage” employers, plus a paid newsletter breaking down each one. Ad-free, reader-funded.
🪟 The Colorado Compliance Window
Colorado’s ban is coming, and other states copy each other fast. That means a 2–4 week scramble where HR departments panic about looking “compliant.” Sell them a dead-simple audit template — “here’s how to prove you don’t use surveillance-wage data.” Boring B2B money, but it’s real and it’s timed.
Example: A 29-year-old in Poland cold-DMs HR managers on LinkedIn offering a €300 one-page compliance checklist + a 20-minute walkthrough. The Colorado bill text is his whole cheat sheet.
💸 The Reverse Recruiter
Job seekers have NO idea what the algorithm secretly thinks their “floor” is. So show them the ceiling instead. Run a quick call where you pull real public salary data (Levels.fyi, Glassdoor) so they walk in and anchor HIGH — instead of letting the robot anchor them low.
Example: A 26-year-old in Brazil books “Know Your Number” 30-minute calls via Calendly at R$150 each, promoted on Instagram. Shows clients the top of their real market range so they never accept the first offer.
🛠️ Follow-Up Actions — do these today, free
| Move | Link |
|---|---|
 Opt out of the biggest data broker |
Acxiom opt-out |
 Kill dozens of brokers at once (paid, worth it) |
Wikipedia: data brokers |
 See your REAL market salary |
Levels.fyi |
 Read the full breakdown |
Pluralistic |
 Track the laws |
Colorado Legislature |
Quick Hits
| If you want to… | Then do this |
|---|---|
 Stop leaking “I’m broke” data |
Opt out of data brokers before job hunting |
 Not get lowballed |
Look up your range on Levels.fyi and name YOUR number first |
 Know who does this |
Start/join a public employer list |
| Understand the whole scam |
Read Doctorow’s piece |
| Make money off it |
Sell the scrub, the list, or the “know your number” call |
They built a machine to find out how little you’ll accept. So stop being cheap to a computer — and go get paid to help everyone else do the same.
!