An AI now grades AI to fool dating apps’ AI | Tinder rented a pickleball court
Tinder rents a pickleball court. The line stretches out the door.
The app that taught a generation to meet without leaving the couch has been quietly hosting in-person events in Los Angeles since March: pickleball, silent discos, dog happy hours. The most recent gathering near Santa Monica State Beach hit capacity. Pickleball has been the fastest-growing sport in the U.S. for three years running, per the Sports & Fitness Industry Association — Tinder did not pick the activity at random; it picked the room where 20-somethings already gather, then offered to pay for the drinks. Some attendees were not Tinder users at all. They came with friends. The flagship metric of a swiping company is now whether the line wraps the block.
The BBC declares dating apps a public-health issue.
BBC Future published a brief treating swiping fatigue as a clinical condition: prescribed pauses, recommended offline alternatives, citations to behavioural research. Researchers describe these engagement loops using the same vocabulary as slot machines — variable-ratio reinforcement, where unpredictable rewards make the behaviour hardest to stop. Hinge built a brand around this insight seven years ago with the tagline “designed to be deleted.” The BBC piece reads like mainstream press finally catching up to what one of the apps already wrote on its own marketing.
Match Group calms investors by promising fewer swipes.
MTCH stock steadied after Q1 earnings, propped up by CEO Spencer Rascoff’s pivot to live events, group dating, and an AI redesign. Tinder alone generates the majority of Match Group revenue — the IRL push is not a side experiment, it is the empire’s main bet. The same playbook that built the swipe economy is now being unwound by the company that depended on it most. The growth thesis for a swipe-revenue business is now to discourage swiping. It is the most polite “we have run out of ideas” any public board has ever sent shareholders.
An app where you can’t ghost, can’t tap “like”, and can’t use it for more than 11 minutes a day.
A fitness-focused dating app called ateam relaunched in New York with an 8,000-person waitlist, a 40-person committee of trainers and “cultural tastemakers”, and a deliberately punishing UX. Daily usage is capped at 11 minutes. The “Like” button has been removed to push users into actual conversation. Ghosting is policy-banned. The product hypothesis is the inverse of every dating app for the last decade: more constraints, fewer matches, longer attention. The wellness category has been pricing access to scarcity for years — soulcycle, gyms, Oura rings — and ateam is the first dating app to apply the same business model to people.
👉 Related: Features 300+ in Dating Pro
Hinge takes its loneliness-grant program to France.
Hinge’s One More Hour fund (now $1.5M global) sends $150–200K to French community groups after Ipsos polling found 62% of Gen Z there report loneliness. France becomes the third country for a programme that started in three U.S. cities in 2023 and has now donated $2.3M globally. The dating app’s central pitch is increasingly that you’d prefer to meet someone outside the dating app. The Marseille “Assemble” event lands May 30, presumably no app required.
A dating-app CEO publicly questions whether India’s low divorce rate is a good thing.
The CEO told NDTV that “suffocating” marriages have normalised suffering — a sentence that probably tested better with the marketing team than with anyone’s mother-in-law. India’s divorce rate sits below 2%, the lowest among large economies. Social scientists have flagged for two decades the gap between “people who stay married” and “people who report being happy in their marriage” — the new note is that a private dating company now publishes the same critique that academia has been making since the early 2000s. The dating industry is in the business of public sociology.
Thursday’s COO says dating apps “had their moment”. Thursday closed its dating app in 2025.
Brenton Bright told a recent interview that 2026 daters are choosing IRL over algorithms. Thursday now operates 200+ event hosts across cities organising weekly singles parties, after shuttering its swipe product last year. The economics are quietly interesting: matchmaker-style margins on a £30–60 ticket beat what any subscription app earns per user per month, but only when a host can fill the venue every week. The new business is selling tickets to the absence of a dating app. The pitch deck must be a thing of poetry.
👉 Built for matchmakers? See Matchmaking Edition
Privacy-first gay dating apps want to dethrone Grindr by being the boring version of Grindr.
Wired profiles a wave of startups offering end-to-end encryption, no ad-tracking, and no location triangulation. The pitch is essentially: same idea, fewer settlements. Grindr was fined €6.5M by Norway’s data authority in 2021 for sharing user data without consent — at the time the largest GDPR penalty against a dating app. The challenger value proposition in this category is the absence of that exact headline. The boring version of Grindr is now the disruptive version.
👉 Related: Free quote for your niche
HundRoses pitches dating profiles as a credit score.
The startup builds “trust signals” into dating profiles using fintech-grade identity verification, targeting markets where catfishing carries serious cultural and legal consequences. Online romance scams cost U.S. consumers a reported $1.3 billion in 2024 according to the FTC — the highest dollar-loss category of consumer fraud in the country. Every HundRoses match now comes with the kind of proof normally requested by a mortgage broker. The romance of online dating is becoming distinctly procedural.
The World Surf League sponsors a dating app called SURF. Naturally.
WSL named SURF — yes, that is the app’s name — as the official singles partner of the US Open of Surfing. The category of niche dating apps tied to a single activity has multiplied to over a thousand worldwide, with the most durable ones (FarmersOnly, Christian Mingle, Bristlr) succeeding for the same structural reason: they restrict the dating pool to a community whose members already meet offline somewhere else. SURF is testing the same principle with a much wetter audience. The branding department had a very simple Monday.
Morning Brew bundled SpaceX’s IPO filing with “dating apps pivot to IRL”.
Both stories ran in the same daily, which is the most Gen Z economic indicator this newsletter has ever produced. The underlying observation is real: every major U.S. dating app has announced an IRL or in-person feature this year — Tinder Events, Hinge One More Hour, Match Group’s group-dating push, Breeze Group Dates, Thursday’s full pivot. When every competitor moves the same direction simultaneously, it is rarely about strategy and usually about a shared admission that the old model has plateaued. The dating-app pivot to live events is the version of that admission with cheaper paperwork than building rockets.
👉 Built for matchmakers? See Matchmaking Edition
Local TV starts calling it “internet self-defense”.
After a string of incidents tied to dating apps in Florida, News4Jax aired a segment where security experts walked viewers through “internet self-defense” routines — share your location, meet in public, reverse-image-search the profile. The phrase “internet self-defense” is borrowed from cybersecurity training literature, where it appeared around 2015 to describe consumer habits against phishing. Local TV applying the same vocabulary to first dates is a small but real category shift: dating-app safety is now positioned next to firewall hygiene. The romance is in the protocols.
Layered absurdity: a service grades AI photos so they can fool dating apps’ AI.
DatePhotos.AI launched the Realness Score, rating each AI-generated profile photo from 1 to 100 across skin texture, lighting coherence, and the telltale AI deformities of irregular hands and teeth. Anything above 85 supposedly survives Tinder’s AI detection. The system is a small case study in what researchers call an adversarial loop: a generator improves to fool a detector, the detector improves to catch the generator, the generator adapts again. There are now three AIs in the room — one making the face, one grading the face, one trying to flag the face — and the only human in the chain is the person looking for a date.