What Is Algorithmic Bias? Why the Internet Keeps Treating You Like You’re 96

By Chris Clark, founder of Senior Tech Cafe

Algorithmic bias is what happens when a computer system that makes choices about people, from which ads you see to whose job application gets read, treats some groups unfairly because of skewed data or lazy assumptions built into it. For older adults, the most common version is being sorted by age instead of by interests.

Have you ever hopped online to buy a simple pair of hiking boots, only to watch your screen fill up with ads for walk-in bathtubs, “final expense” insurance, and medications with side effects scarier than whatever they’re supposed to treat?

You didn’t type “I feel ancient today” into Google. You wanted boots. But the internet took one look at you and decided your only remaining hobbies are soaking in specialized tubs and contemplating the void.

Your computer isn’t broken. You’ve just walked into an invisible wall, and that wall has a name: algorithmic bias.

Think of an algorithm as a fast, over-caffeinated digital butler. Its whole job is to guess what you want before you ask. But for older adults, this butler is often working from a terrible set of instructions, and the result is an internet that feels like it was built for somebody else.

The Negative Language Trap

AI learns to “think” by reading staggering amounts of text from the internet. And the internet, bless its heart, is not always a kind place. Over the years, these systems have soaked up plenty of lazy, negative assumptions about aging.

So when a system runs into words it associates with older adults, it can shift its behavior. It might surface fewer interesting job listings, explain things to you like you’re five, or bury the travel articles under a pile of medical news. Nobody programmed it to be condescending. It just learned condescension by reading ours.

AI Bias Examples That Hit Older Adults

The best-known AI bias examples aren’t about annoying ads. They’re about jobs and money, where an automated “no” costs you something real. Three cases show how it happens:

• The hiring software that screened out older applicants. In 2023 an online tutoring company, iTutorGroup, paid $365,000 to settle a federal age-discrimination case. The government said its hiring software automatically rejected women 55 and older and men 60 and older, turning away more than 200 qualified applicants. The company denied wrongdoing.
• The job ads you never saw. Facebook used to let advertisers aim job ads at younger users only. In 2019, settling lawsuits from civil rights groups, it agreed to stop allowing age, gender and zip-code targeting on job, housing and credit ads.
• The résumé reader that learned the wrong lesson. Amazon dropped an experimental hiring tool, Reuters reported in 2018, after finding it marked down résumés that included the word “women’s,” as in “women’s chess club.” It had taught itself from ten years of résumés, most of them from men.

The pattern is the same each time. The system copies whatever bias sits in its instructions or its history, then repeats it at a speed no human hiring manager could match.

The Hidden Filters: Where You’ll Hit These Walls

These biases do more than serve up annoying ads. They reshape your whole internet, sitting between you and the good stuff. Here’s where the walls tend to pop up.

Your Shrinking Newsfeed (the Filter Bubble)

A filter bubble is the narrow slice of the internet you end up seeing when a feed only shows you more of what it thinks you already like. Social media platforms want you clicking, full stop. If the algorithm decides seniors only care about health scares, nostalgia and scam warnings, then health scares, nostalgia and scam warnings are what you’ll get.

Do this long enough and your feed becomes a tiny room. The new gadgets, the local events, the travel daydreams slip behind the fence. Your digital world shrinks, one assumption at a time. And a feed that only shows you one kind of story is also where fake news spreads best, because nothing in the room argues with it.

Pricing You Can’t Quite See

Here’s one worth knowing about. Some companies use “dynamic pricing,” where the cost of something shifts with demand, timing, location or inventory. That part isn’t aimed at you personally. Everybody shopping at that moment sees it.

“Personalized pricing” is different: a company uses your personal data to estimate how much you’d be willing to pay. Regulators are starting to poke at this. New York now requires businesses that use your personal data to set a price to tell you so, and a federal judge upheld that law in 2025.

Everywhere else, it’s usually impossible to know why you’re seeing a particular price, which is the real problem. So before you buy something big, compare the price in a second browser, on another device, or straight from the retailer’s own website.

The High-Stakes Stuff

The walls get more serious in places like credit, insurance, housing and hiring, where automated systems increasingly help make the call. If the data or the assumptions behind those systems are flawed, they can spit out unfair results, and unlike a bad ad, a bad loan decision actually costs you.

That’s why these areas draw the most legal scrutiny. If you’re turned down for credit, the lender has to give you the specific reasons, or tell you how to ask for them. If an automated “no” on a loan or a job ever feels off, ask for the reason in writing.

Tearing Down the Wall: How to Retrain the Algorithm

Now the good news: you’re not stuck. Algorithms are stubborn, but they’re also easily retrained. These habits do the most, roughly in order:

• Use the “not interested” buttons. When a platform shows you junk, tap “Hide ad,” “Not interested” or “See less like this.” You’re handing the butler a direct correction, which works far better than hoping it figures things out on its own.
• Tell it what you do want. Search for and follow the things you actually care about: travel, gardening, photography, recipes, local events, that band you’ve loved since 1971. Good clues in, better feed out.
• Turn off personalized ads. Google and Facebook both let you limit ads built from your profile. You’ll still see ads, but they’ll be chosen more by what’s on the page than by guesses about you. Here’s how to turn off personalized ads step by step.
• Clear your cookies now and then. This wipes some of what sites remember about you. It’s not a magic eraser: logins, accounts and device IDs can still tag along, and some website trackers don’t use cookies at all. Think of it as a light disguise, not witness protection.
• Shop and research in a private window. For price checks or health questions, a private window keeps that session from piling onto your browsing profile, as long as you don’t sign in.

You don’t have to turn your browsing history into a teenager’s midlife crisis to get treated like an individual. A few small habits will do it.

So What Happens Next?

The rulebook is still being written, and slowly. In 2022 the White House published a “Blueprint for an AI Bill of Rights” that named unfair automated decisions, age included, as a real concern. It was guidance, not law, and it was taken down from the White House website in January 2025.

What’s left is the older laws that never went away. Age discrimination in hiring (for anyone 40 or older) and in credit is still illegal when a computer makes the call, plus a patchwork of state rules like New York’s pricing law. Until the official rulebook catches up, your own awareness is the best tool you’ve got.

The internet belongs to you just as much as it belongs to the teenagers of Silicon Valley. You now know the walls are there, and a wall you can see is a wall you can walk around. For more plain-English guides to making the internet behave, start with our online safety guide for seniors. No jargon, no lectures, no bathtub ads.

Questions About Algorithmic Bias

What Exactly Is an Algorithm?

An algorithm is a set of rules a computer follows to solve a problem or make a choice. Online, the important ones study what you click, watch and buy to guess what you’ll want next, then decide what to show you. They’re confident. They’re fast. And they’re frequently wrong, especially when they lean on your age.

Why Do I Get so Many Ads for Medical Alert Buttons and Final Expense Insurance?

Because many advertisers aim at broad age brackets and assume every older adult wants the same handful of products. Once you’re in that bracket, your actual hobbies barely get a vote. Telling the platform you’re not interested, following things you do care about, and turning off ad personalization all help break the pattern.

What Is a Filter Bubble?

A filter bubble is the narrow slice of the internet you end up seeing when a feed only shows you more of what it thinks you already like. Each click teaches it to show you more of the same, and everything else slips out of view. Following a wider range of topics and sources helps pop it.

Is Algorithmic Bias Against the Law?

Sometimes. If an automated system turns you down for a job because you’re 40 or older, or for credit because of your age, existing federal laws against age discrimination still apply, even though a computer made the call. Federal fair-housing law doesn’t list age, though some states do. Stereotyped ads on social media are legal, just annoying.