Auto-swiping: how it ends, and why we don't build it

The pitch is tempting: let software do the liking while you get on with your day, then you just sort through the matches. We get asked about it every few months — "can your service do that?" Here is the long answer, because the short one is not "no, that's cheating" but "here is what will actually happen to you."

The short version. Auto-swiping gets detected, costs you your account, and with face verification at signup a ban is now close to permanent. But the bigger problem is simpler: mass-liking damages your visibility inside the app and brings matches that never turn into conversations. We don't do it, and not out of moralising — it just doesn't solve the problem people come with.

Four different things called auto-swiping

Scripts and extensions. The basic version: a program presses "like" for you. These have been sitting on GitHub for years, install into a browser or a phone, and work right up until the first check.

Paid bot services. One level up: the service doesn't just like, it runs the conversation, sets up the date and hands you a phone number. CupidBot, the best known of them, promised men exactly that.

AI agents. The fashion of the last two years: you tell an agent who you are looking for, and it walks around the app on your behalf. Technically the same bot, only chattier.

Features built by the apps themselves. This one is different. Tinder showed AI-assisted matching in spring 2026, Bumble is building an in-app assistant meant to replace endless swiping. When the automation comes from the platform, it is legal by definition — their product, their rules.

The distinction matters. The apps are not against automation as such. They are against someone else's software walking into their service and stepping around their defences.

What happens if you try

You get caught by rhythm. A human swipes unevenly: fast, then a pause, then distracted by something. A script produces clean intervals. Add the obvious tells — a like in under half a second, a hundred percent right swipes, thousands of actions in a row with no gap for sleep.

A shadowban is worse than a normal one. The app keeps working. You swipe, you write, you like. You are simply not shown to anyone any more. Matches stop, likes go nowhere, and it can take a long while to work out what happened.

Starting a fresh account is no longer an option. This is the real change of the past year. Tinder made face verification mandatory: at signup you record a short video selfie, which the system checks against the photos in your profile. It rolled out across the US in autumn 2025, became mandatory for new users in the UK in March 2026, and is being extended to people who registered years ago. The same check flags attempts to run several accounts on one face. In plain terms: a ban used to mean "sign up again", now it means "this app is closed to you."

For context, the company says the check cut users' exposure to scams and spam by more than sixty percent, and that fake accounts and spam accounted for ninety-eight percent of all its moderation actions in 2025. The apps are not fighting you personally, but anyone who looks like a bot gets caught in the same net.

And there are lawsuits. In March 2024 Match Group sued CupidBot, arguing the service circumvented Tinder's protections and interfered with how the platform works, with claims including the Computer Fraud and Abuse Act. For a customer that means something very practical: the service you handed your account credentials to can vanish overnight, along with your money.

Does it even work?

Suppose you don't get caught. Does liking everyone actually produce results?

No, and you can see why from how the apps are built. The algorithm looks not only at how many likes you receive but at how selective you are. Liking everything reads either as a bot or as desperation, and your visibility drops afterwards. Services that analyse exported swipe data report that selective users match several times more often than mass swipers. The exact multiplier can't be verified, but every source points the same way.

Then comes the part nobody plans for. Say the matches do pile up. What do you do with them? The person on the other side liked your profile, not a bot. They expect a conversation, and the conversation is yours to have. If the matches were collected blindly, you end up with a hundred people you have nothing in common with and not one reason to write.

And the third effect is the ugliest. Auto-swiping degrades other people's feeds too. When a woman receives a flood of matches from men who liked her without looking, half of those conversations will never start. She gets noise instead of attention, and answers even less often next time. It works as a vicious circle: the more bots, the lower the response rate for everyone real.

What daters themselves think

Here is the most telling figure of the year. A Match Group survey in June 2026, a thousand Americans aged 18 to 39: almost half view AI in a romantic context negatively. In the same survey three quarters use ChatGPT, and roughly two thirds say AI helped them build a better profile, open a conversation and keep it going.

Coffee Meets Bagel found the same split even more sharply: around eighty percent are comfortable asking AI for help with their own profile, yet most would lose interest in a match who had done exactly that.

People are not against tools. People are against being talked to by a program. The line sits precisely where help turns into substitution: helping you phrase something is fine, sending it for you and booking a date without you is a lie.

Why we don't build it

Three reasons, none of them about morality.

You pay with your account. The cost of getting caught used to be small — banned, register again. Face verification closed that door. We are not going to ship a tool that has a decent chance of cutting someone off from the app where they are trying to meet people.

It doesn't solve the problem. Nobody arrives with "I need to send more likes." They arrive with "why is nobody responding to me?" More likes on the same profile produce more silence, not more dates. We would rather work on the cause: Profile Review goes through the whole profile and says what is not landing — an AI read in a couple of minutes, then written feedback from real people in the audience you want to attract.

You should be chosen for you. The whole point of what we build is that the profile shows a real person, clearly and honestly. A match won by a bot loses its value in the first minute of the conversation, because the conversation is yours.

The conditions under which we would build it

Never say never. There are three conditions under which automation in dating stops being cheating, and the apps are moving that way themselves.

An official interface. If the platform lets you plug in an assistant, the way Tinder and Bumble have started doing with their own AI features, the question of getting caught disappears.

Visibility for the other person. If it is clear that an assistant helped draft the first message, nobody is being deceived. The surveys above say exactly this: people object to the concealment, not to the tool.

Help with choosing, not faked activity. "Like everyone for me" and "show me only the people I'd actually have something with" are different products. The first breaks the app; the second saves time on both sides.

If those three line up, we'll build it: not a bot pretending to be you, but an assistant that protects your attention.

What to do instead

Automation promises to take over the volume. The problem is almost always quality, and it gets fixed in four places.

• Swipe selectively. Not only good manners: selectivity improves how the algorithm treats you. How the apps decide who sees you is covered separately in how the dating app algorithm works.
• Start with the photos. The decision about you is made in a second, and it is made on the first frame. Photo Selection scores every shot on gaze, light, framing and background, ranks them by strength and explains each verdict, including the ones dragging the profile down.
• Rewrite the text about yourself. "I love travelling and good company" gives the reader nothing to hold on to. Bio Builder asks questions and turns your answers into a text that sounds like you rather than like a template.
• See the profile through someone else's eyes. You cannot look at your own photo the way a stranger does. Profile Review gives you both readings: the AI breakdown and written feedback from real people. If you don't know where to start, begin with what's wrong with my dating profile.
• Write first yourself, but prepared. Help with phrasing is fine; blasting the same line at everyone is a bot with extra steps. Message Helper works only with what you bring it: their profile, or a conversation that went quiet.

The difference between this and a bot is simple. A bot fakes your activity. Everything above gives people a reason to pick you. The apps punish the first and promote the second.

Auto-swiping promises to skip the queue. What it actually does is move you to the back of it, and you find out late.