Tinder’s ‘Chemistry’ Experiment Uses AI (and Your Photos) to Find Better Matches

Tinder’s parent company, Match Group, is betting big on artificial intelligence to help its dating app bounce back after nine straight quarters of declining paying subscribers. The company announced during its Q3 earnings call that Tinder is testing a new AI-driven feature called Chemistry, designed to understand users better and improve match recommendations.

A New Kind of Matchmaking

Chemistry will ask users a series of questions and, with their permission, access their Camera Roll to learn more about their interests and personality. The feature is already being tested in New Zealand and Australia, and Match says it will be a “major pillar” of Tinder’s 2026 product experience.

Match Group CEO Spencer Rascoff said the goal is to make Tinder feel more personalized, not just transactional. Essentially, the AI might notice, for example, that you have several hiking photos and pair you with someone who also enjoys the outdoors.

This experiment mirrors a recent move by Meta, which began asking users for permission to scan phone photos that haven’t been uploaded yet to offer AI-powered suggestions and edits. But in both cases, the benefits for users are debatable, especially as privacy concerns linger around how personal data and photos are analyzed.

Testing Comes at a Cost

The rollout of Chemistry isn’t cheap. Match said Tinder’s direct revenue will take a $14 million hit in Q4 due to the ongoing product tests. Combined with weaker spending trends across the dating industry, Match revised its Q4 revenue guidance to between $865 million and $875 million—below Wall Street’s expected $884 million.

Tinder’s revenue in Q3 dropped 3% year-over-year, and paying users fell 7%. That’s a worrying trend for a platform that still dominates the dating space but faces growing competition from both niche apps and a shift toward offline, real-world interactions.

AI Across Tinder’s Ecosystem

Beyond matchmaking, Tinder is using AI in other parts of the app. Its large language model (LLM)-based system can flag potentially offensive messages before they’re sent, prompting users with a quick “Are you sure?” The app also helps users select their most appealing photos—something many daters may quietly appreciate.

These AI additions show Tinder’s broader push to modernize its experience and re-engage users who’ve grown tired of swiping fatigue. Still, they come at a time when the app must walk a fine line between personalization and privacy.

A Market That’s Falling Out of Love

Even with product redesigns, new “modes,” double-date features, and revamped profiles, Tinder is navigating a tough market. Many younger users are spending less time (and money) on dating apps, and U.S. consumers are tightening their wallets as economic uncertainty lingers.

Match’s overall Q3 performance was roughly in line with analyst expectations: $914.2 million in revenue (up 2% year-over-year) and 62 cents in earnings per share, just shy of forecasts. But the decline in Tinder’s core metrics underscores the challenge ahead.

The real question is whether AI-driven “Chemistry” can make Tinder feel fresh again—or if users have simply swiped past the era of digital dating.

Via