Tinder’s 50 million users are burning out. The app is betting AI can fix what swiping broke

The timing isn’t incidental. Dating apps are facing a user crisis. Many report feeling burnt out from endless swiping without meaningful connections. Bots have proliferated. Younger users are spending less time on the platforms. For Tinder, which claims 50 million monthly users, AI represents both a technical evolution and an existential necessity.

CEO Spencer Rascoff at Tinder’s first-ever product launch event in Los Angeles on March 12, 2026. (Image credit: Anuj Bhatia/Indian Express)

‘Using AI to remove friction’

“We are using AI to help bring people together… We look at the entire user journey and try to identify the possible friction points and opportunities. Based on those challenges, we think about the ways to fix them. If AI is one of those ways, then we will use it,” Mark Kantor, Head of Product at Tinder, told indianexpress.com in an interview.

“During onboarding, there’s often friction — questions like ‘How do I create a profile?’ or ‘What do I do?’ So we use AI to help remove some of that friction and help people express themselves more authentically,” he explained, citing the AI photo selection feature which can look through tens of thousands of your photos and suggest a few that Tinder thinks will perform well.

Mark Kantor, Head of Product at Tinder. (Image: Anuj Bhatia/The Indian Express)

“We use AI in our recommendation algorithm to help figure out who you might be interested in and show you the right person at the right time.” He said this also had a safety perspective: to make sure that the people on the platform are real and have only one profile. “For us, it’s really about figuring out how we can use this amazing technology to remove friction, make people more expressive, and make the platform more intelligent and safer.”

Kantor, who heads cross-functional teams across product, design, revenue, trust and safety, and research, said the entire organisation has lately become more product-focused, much like a tech startup. That shift also reflects in how Tinder is able to release new features at a much faster pace. AI, he said, plays a big role in how the company has adopted this approach.

“On the development side, more than half of our code is actually written by generative AI,” he explained. “We are able to prototype faster and move into development more quickly.” For instance, they make small teams spend time together on an idea and come up with a working prototype in a couple of hours. We then take that prototype, test it internally, build confidence, and once we decide to build it, move from prototype to production without having to write 50-page specs or create Figmas with 300 screens. That approach dramatically improves speed.”

A move away from the ‘swipe model’

Tinder, perhaps the best-known matchmaking app, is pivoting toward AI to reconnect with younger daters worn down by low-quality matches and an explosion of bots that make real connections harder to find. The West Hollywood company is responding with two features designed to replace volume with precision.

The first is Chemistry — live in Australia and New Zealand, coming soon to the US and Canada — which analyses profile information, question responses, and photo insights to surface highly compatible matches. Rather than a free-form chatbot, it uses structured multiple-choice prompts. “We ‘spoon-feed’ the questions,” said Claire Watanabe, VP of Product, “asking things like, ‘Are you more of an evening person or a morning person?’ It’s less effortful and more fun, while still capturing the insights you’d get from an actual conversation.” Hillary Paine, SVP of Product, frames it as a direct counter to swipe fatigue: “It’s an intentional decision to say, here is your highest-quality, most compatible match. You could stop there.”

Hillary Paine, SVP, Product (left), and Claire Watanabe, VP Product at Tinder (right) (Image: Anuj Bhatia/The Indian Express)

The second is Learning Mode, a real-time recommendation system that builds a preference profile from your swipe behaviour and — crucially — makes that learning visible to users. “You keep swiping right on surfers. We give that feedback back to you,” Kantor added. Internal testing found women who used it were more likely to return to the app within the first week.

Both features point to the same diagnosis: the swipe model that Tinder invented is showing its age.

Dating under pressure

This shift poses a threat to dating apps like Tinder, but the platform is hoping that its new changes will bring attention in an increasingly crowded market and introduce features that bring younger users back to the app.

“Younger daters are more interested in spending their time intentionally, for example. They are looking for more efficient outcomes, more social and community-forward experiences, and lower-pressure dynamics. Those may be Gen Z trends, but they are also broader dating trends that we are innovating for,” Paine said. According to Paine, almost 60 per cent of Tinder’s user base is under 30.

Not only has Tinder refreshed the app’s user interface, giving it a transparent aesthetic, similar to Apple’s Liquid Glass UI, but many of its new features are also designed with modern, younger daters in mind. One such feature is a real-time video speed-dating experience within the app, set to arrive later this spring, where users who have undergone photo verification will be able to join scheduled virtual events for three-minute video chats. There is also a new “Events” feature, which capitalises on the growing in-person dating scene, allowing users to discover local happenings and see who else is interested in attending. The feature will debut first in Los Angeles as part of a pilot run.

‘AI address safety of users’

But one area where AI is making the biggest impact is user safety on dating apps, a major concern for women daters and those who belong to the LGBTQI+ community on Tinder.

“I think AI can be positive in two big ways. The first is through some of the features where we use AI to try to preemptively shape behaviour. For example, features like ‘Are You Sure?’ allow us to use AI to detect if a message might be abusive or inappropriate and prompt a user to reconsider it. The second area is helping us proactively detect behaviour that breaks our rules. About 80 per cent of the moderation actions we take right now happen without the need for a user report. I want that number to be as close to 100 per cent as possible, and the key to getting there is AI,” said Yoel Roth, SVP of Trust & Safety at Tinder.

Tinder also plans to add an auto-blur feature to its “Does This Bother You?” tool, which detects potentially inappropriate messages.

India, one of the most populous countries, happens to be a market where Tinder is particularly focused on the safety of its users. “I think we are still at the beginning of our journey in India. The opportunity here is huge. It’s a big country, very connected, and, especially as norms around relationships and marriage evolve, we want Tinder to remain widely available in India. Women’s safety is essential, and we have built and launched in India with women’s safety at the core from the very beginning,” claims Roth.

Yoel Roth, SVP, Trust & Safety at Tinder. (Image: Anuj Bhatia/The Indian Express)

The AI technology that Tinder is adding to monitor harassment and gender-based violence sends a powerful message: AI belongs in every corner of our lives. “Before the call starts, to use the video speed-dating feature, you need to be a Tinder user in good standing. But let’s say you are on a video speed date and place your phone inappropriately. In that situation, we have AI running on the device during the call that can detect things like nudity. If that happens, we would end the call and enforce our rules,” explains Roth.

“I don’t really believe in monetising safety features. I think safety is a right, not a privilege. More than being a right, safety is the foundation of product success. Being a paying subscriber does not mean you are exempt from our rules,” he said, when asked if Tinder plans to monetise safety features.

Tinder is aware that AI is a useful tool, but not every feature needs to be AI-powered.

“On the user-facing side, we don’t go in saying, ‘We need eight more AI features.’ It’s really about what people need. For example, for IRL events, we used AI to help build and prototype the feature, but it’s not an AI-powered feature itself. We think IRL addresses what people really need. For us, it’s about figuring out what helps people connect. If AI can help us do it, great. If not, that’s equally fine,” Kantor explains.

From Swipe to the ‘AI matchmaker’

Online dating has gone through many evolutions. Starting in the mid-’90s, platforms were largely profile-based, matching users with shared interests and values. In the early 2010s, the launch of Tinder introduced the swipe model. Other apps, such as Grindr, which caters to gay men, use a location-based model, allowing users to browse potential dates in their area. Newer dating apps, however, are trying to give users a break from swipe fatigue.

Tinder wants to be part of the modern dating scene , especially among young singles, but it can only do so by embracing technology and keeping up with new dating trends. That is where AI comes in, with Tinder becoming the AI matchmaker.

“I think we are just trying to evolve the experience and add even more meaning to it. If we go back six months, what you were swiping on were just a couple of photos of a person. We have seen that as we add more depth to a profile, even surfacing interests that were previously buried, it changes the experience. You’re still swiping, but you are not just swiping on two photos — you are swiping on a more well-rounded version of a person. The swipe itself is a great gesture: super intuitive and very easy to do. We just wanted to make it mean more,” Kantor responded when asked if swipe will be the part of the core app experience in the next five years.

“I think it’s still about connecting single people. What has changed, to me, is the definition of success. For some people, it’s a one-on-one dinner date; for others, it’s a group pickleball game. As long as we are connecting people who are seeking connection, then we have done our job,” Kantor explains.