Here's an unsettling thought to sit with the next time you shop online: the price you see may not be the price the person next to you sees. Same product, same moment, same website — different number, because an AI has studied everything it knows about you and decided exactly how much you, specifically, are willing to pay. It's not science fiction and it's not a glitch. It has a name — surveillance pricing — and it is quietly pulling money out of your pocket right now.
This belongs on a site about AI and money because it's the purest example of AI touching your money without your permission. Most of what I write here is about using AI to earn more. This is the flip side: AI being used, by companies you buy from, to make you pay more. And the sums aren't trivial — one investigation found surveillance pricing could cost some shoppers an extra $1,200 a year on groceries alone, before you even get to rent, travel, and everything else. So let me show you how it actually works, what it's costing you across your whole life, who's fighting it, and — most usefully — how to pay less.
What surveillance pricing actually is
First, an important distinction, because not all price changes are sinister. Dynamic pricing — where a price moves with the market — is old and mostly fair. Airline seats cost more near a holiday, hotel rooms cost more during a big event, surge pricing kicks in when demand spikes. That responds to supply and demand, and it hits everyone the same way at a given moment.
Surveillance pricing is a different animal entirely. It sets a price based on you as an individual — your data, your behavior, your perceived willingness to pay — rather than the market as a whole. Regulators draw exactly this line: dynamic pricing reacts to inventory and demand; surveillance pricing reacts to characteristics of the specific consumer. That's the shift AI made possible. Older systems couldn't model you personally at scale; modern machine learning can ingest thousands of signals about you and output a bespoke price in milliseconds. The result is a market where the sticker isn't a fact about the product — it's a guess about you.
It's already here, and it's measurable
This isn't a hypothetical future risk. Regulators and journalists have already caught it in the wild, and the specifics are jarring:
| Case | What was found |
|---|---|
| Instacart probe | Surveillance pricing could lead to roughly $1,200 more per year in grocery spend for some consumers |
| FTC issue spotlight | Signals as granular as mouse movements or items abandoned in a cart could feed a tailored price |
| RealPage (rentals) | DOJ sued over software alleged to enable algorithmic rent price-fixing among competing landlords; settled in 2026 |
Read the middle row again, because it's the one that should change how you feel about this. The FTC found that a company could adjust your price based on how your mouse moves across the screen, or the fact that you left something in your cart — reading hesitation, desire, and urgency in real time and pricing accordingly. That's not pricing a product; that's pricing your psychology. And the RealPage case shows it isn't just retail: the software that helped set rents for millions of apartments drew a federal antitrust suit over algorithmic coordination. When the algorithm decides your grocery bill and your rent, this stops being a curiosity and becomes a line item in your life.
How the AI decides what to charge you
So what is the machine actually looking at when it prices you? Far more than you'd guess. The signals fall into a few buckets, and together they build a startlingly accurate model of how much you'll tolerate:
| Signal type | What it tells the algorithm |
| Device & OS | Shopping on a premium phone can flag you as less price-sensitive |
| Location | Your neighborhood proxies for income and local competition |
| History | What you've bought, browsed, and paid before — your personal ceiling |
| Behavior in the moment | Urgency, hesitation, cart-abandonment, time of day |
| Loyalty & logins | Ironically, being a loyal logged-in customer can mean worse prices |
Notice the cruel twist in that last row. The signals that feel like convenience — staying logged in, using the app, being a loyal repeat buyer — are exactly what let the system profile you and, sometimes, charge you more than a cold anonymous visitor. The better a company knows you, the more precisely it can find your personal maximum. This is the same "your data is the product" dynamic behind so much of the AI economy, the flip side of the value extraction I described in how the AI wealth boom locks ordinary people out — except here the extraction happens one checkout at a time.
The money math: what it's really costing you
Let's add it up honestly, because scattered across categories this hides in plain sight. No single overcharge is big enough to notice, which is precisely why it works. Here's a plausible annual hit for a household that shops and rents like most people:
| Category | Estimated surveillance-pricing cost / year |
| Groceries & delivery | Up to ~$1,200 (per the Instacart finding) |
| Rent (algorithmic pricing markets) | Hundreds to ~$1,000+ |
| Travel & airfare (personalized fares) | ~$100–400 |
| Retail & subscriptions | ~$100–300 |
| Plausible total | ~$1,500–3,000+ per year |
Even at the low end, that's real money — a vacation, an emergency fund contribution, a month of groceries — quietly redirected from your budget to a retailer's margin because an algorithm decided you'd pay it without noticing. And the genius (for them) is the invisibility: you never see the other price, so you never feel robbed. You just pay a little more, everywhere, forever. That's why I think of surveillance pricing as a tax you didn't vote for — small per transaction, significant per year, and completely opaque.
The fight to rein it in
The good news is that regulators and lawmakers have noticed, and 2026 has been a turning point. Here's where the pushback stands:
| Front | Action |
| FTC | Ongoing surveillance-pricing work; weighing whether personalized, data-driven pricing needs mandatory disclosure |
| DOJ | Sued RealPage over algorithmic rent coordination; settled in 2026 |
| States | 24 states introduced 51 bills curbing algorithmic pricing in early 2025 — up from 10 the year before |
| New York | Proposed act would ban personalized algorithmic pricing, with statutory damages of at least $5,000 per violation |
That jump from 10 bills to 51 in a single year tells you how fast this went from fringe concern to political priority. The core demand across all of it is transparency — the right to know when a price was set by a machine studying you, rather than by the market. Whether that becomes law everywhere is the open question, and it sits inside the bigger debate I explored in does AI regulation help or block progress: regulators are trying to curb a genuine harm without banning ordinary, fair pricing. For now, the rules are a patchwork, which means the responsibility to protect yourself still mostly falls on you.
How to protect yourself and pay less
Here's the practical part, because you're not powerless. Surveillance pricing runs on data and profiling, so the counter-move is to give it less to work with and to break its assumptions. None of this is exotic — it's a handful of habits that can claw back real money:
Shop like a stranger. Browse in a private/incognito window, logged out, so the site can't tie the price to your profile. Compare that price to what you see logged in — the gap is your surveillance premium.
Check across devices. Look at the same product on your phone and a cheap laptop, and on different networks. If the prices differ, you've caught the algorithm profiling your hardware, and you can buy on whichever shows the lower number.
Don't broadcast urgency. Leaving items in your cart, returning to the same listing repeatedly, or searching a flight over and over can signal desperation and raise your price. Clear cookies between sessions, and don't let the system watch you circle.
Use the machines against the machines. Price-tracking tools and comparison AIs work for you, checking history and rivals so you buy at genuine lows. This is where the agentic-shopping tools I wrote about could actually flip the power back to consumers — an AI negotiating on your behalf against the AI pricing you. You can even ask a general assistant like ChatGPT, Claude, or Gemini to sanity-check whether a "deal" really is one.
Be a harder target for travel. For flights and hotels — the most aggressively personalized category, as I noted in the AI travel-booking shakeup — compare incognito, check a couple of sources, and avoid obsessively reloading the same fare.
Done consistently, these habits can recover a meaningful chunk of that $1,500–3,000 a year. It's not about paranoia; it's about not being the easiest wallet in the room.
Who profits — and the honest other side
Follow the money and you see why this is spreading fast: surveillance pricing is enormously profitable for the businesses that deploy it, and a booming market for the vendors who sell the pricing AI. Every dollar of "personalized" markup drops almost straight to margin. That's a powerful incentive, and it's why so many retailers, delivery apps, and landlords quietly adopted these tools — the same margin-capture logic that makes AI attractive to every enterprise, seen from the customer's losing side.
But I'll be fair about the nuance, because it matters. Not all personalized pricing is evil — a targeted discount is personalized pricing, and getting a lower price because you're a loyal customer is the friendly version. The line between "a deal just for you" and "a penalty just for you" is genuinely blurry, which is exactly why regulators are struggling to write clean rules. The problem isn't that prices vary; it's that they vary secretly, based on surveillance you never consented to, in a direction that usually favors the seller. Transparency — knowing when and why — is what separates a fair market from a rigged one.
What this means for you
Depending on who you are, here's the read.
If you're a shopper — which is everyone — the takeaway is simple: the price is no longer a fixed fact, so stop treating it like one. Build the incognito-compare habit, resist manufactured urgency, and use price tools. You'll pay less, and you'll stop being the profiled wallet the algorithm counts on.
If you run a business, this is a fork with real stakes. Surveillance pricing can juice margins short-term, but the regulatory and reputational risk is rising fast — 51 state bills and a $5,000-per-violation proposal are not noise. Building trust with transparent, fair pricing is looking less like idealism and more like risk management, especially as the backlash grows.
If you invest or watch the industry, understand that pricing AI is a real, fast-growing category — and also a regulatory target with a bullseye on it. The same tension I flagged in the AI bubble math applies: a technology can be lucrative today and legally constrained tomorrow. And the erosion of trust it causes feeds directly into the broader crisis of trust AI is creating — a cost that doesn't show up on any one company's balance sheet but is very real for society.
The honest take
What bothers me most about surveillance pricing isn't any single overcharge — it's the quiet asymmetry of it. The seller knows everything about you; you know almost nothing about how your price was set. That imbalance, multiplied across every purchase, is how a few dollars here and there becomes thousands a year flowing from ordinary people to whoever owns the pricing algorithm. It's the same story I keep telling on this site from different angles: AI is a lever, and right now it mostly amplifies the power of whoever already holds it — here, the seller across the counter from you.
But the counter-lesson is just as real, and more hopeful. The same technology that prices you can defend you. An AI that shops, compares, and negotiates on your behalf turns the fight into machine-versus-machine, and that's a fight you can actually win — the way I described in how AI agents are starting to buy on your behalf and how the best defense against AI-powered scams is AI-powered awareness. The people who lose to surveillance pricing are the ones who don't know it exists. Now you do.
So here's the question worth carrying to your next checkout: is the price you're looking at a fact about the product — or a guess about you? Once you start asking that, you stop being the easy wallet, and you start paying what things actually cost.
