Dating apps have been losing people for a while now, and the companies behind them know it. Revenue is down, paying subscribers are leaving, and the users who remain seem tired of the whole thing. So the response from the largest platforms has been predictable: spend more money on artificial intelligence, build new features, and hope that better algorithms will bring people back. Match Group alone has set aside $60 million for AI and product rollouts at Tinder, according to its Q4 2025 earnings disclosures reported by CNBC. That is a serious budget for a problem that may not be a technology problem at all.
The premise behind most of these updates is that smarter software will lead to better matches, and better matches will keep users around longer. It sounds reasonable on paper. But a Forbes Health and OnePoll survey found that 78% of respondents said they felt burned out on dating apps entirely. When nearly 4 out of 5 people using your product say they are exhausted by it, a new AI feature is a strange thing to pin your hopes on.
This article looks at what the major platforms are actually building, where the money is going, and how likely any of it is to fix what has been breaking for years.
The Numbers Behind the Fatigue
Bumble reported Q4 2025 total revenue of $224.2 million, a 14.3% drop compared to the same period the year before. Its total paying users fell 20.5% to 3.3 million. Tinder saw its own revenue decline by 3% year over year, and its paying user count dropped 8% to 8.77 million. These are the 2 largest names in the space, and both are contracting.
The fatigue survey from Forbes Health adds useful context here. Among Millennials and Gen Z specifically, 79% reported exhaustion with online dating. These are the age groups that dating apps depend on most heavily for growth, and they are the ones pulling away fastest.
Hinge is the exception worth noting. Its direct revenue grew 26% year over year to $186.5 million in Q4 2025, making it the strongest performer in the Match Group portfolio. Hinge has positioned itself around relationship intent, and that focus appears to be working better than the broad, high-volume approach that Tinder pioneered.
Relationship Preferences and the Apps That Try to Keep Up
People use dating apps for very different reasons, and the platforms have never been great at accounting for that. Someone looking for a long-term partner and someone trying to find a sugar daddy are using the same handful of apps, often with the same generic matching tools. As apps pour money into AI features heading into 2026, the question is how well they can actually sort through what users want when those wants vary so widely.
A Forbes Health and OnePoll survey found 78% of respondents reported dating app burnout, and 79% of Millennials and Gen Z said they felt exhausted by the process. More technology may not fix a problem rooted in mismatched expectations between users and the platforms they rely on.
What Tinder Is Building With $60 Million
Tinder’s biggest bet is a feature called Chemistry. It uses AI to curate a smaller set of daily matches instead of presenting the endless scroll that the app became known for. The idea is to reduce swipe fatigue by giving users fewer but higher quality suggestions each day. It started rolling out in the U.S. and Canada in late 2025.
There is also Face Check, a safety feature that requires users to submit a video selfie for verification. Tinder’s early internal data showed that verified profiles resulted in over 60% lower exposure to bad actors on the platform, according to a Match Group press release distributed through PRNewswire in October 2025. That is a measurable improvement, and safety has been a persistent complaint from users for years.
The question with Chemistry is harder to answer. Fewer matches per day sounds appealing in theory, but the algorithm still has to be good enough to surface people you actually want to talk to. If it fails at that, showing you 5 bad matches instead of 50 does not accomplish much.
Can AI Fix a Human Problem?
Research from late 2025, as noted in reporting by TechCrunch and What’s Trending, found that many singles want less technology involved in their romantic lives, not more. That puts the entire AI investment thesis in a strange position. The platforms are spending tens of millions building smarter recommendation engines, and a portion of their own users are saying they would prefer the opposite direction.
There is a reasonable case that AI can help with some things. Screening for safety, verifying identities, and filtering out spam accounts are all tasks where automation adds real value. Nobody wants to sort through fake profiles manually.
But match quality is a different matter. Human attraction is inconsistent. People often do not know what they want until they find it, and they routinely contradict their own stated preferences. Training an algorithm on profile data and behavioral signals can improve surface-level compatibility, but it cannot account for chemistry in the way people actually feel it.
Where the Money Tells the Real Story
Hinge’s 26% revenue growth happened without the kind of massive AI spending that Tinder is undertaking. It grew because it attracted users who had a specific goal and then built tools around that goal. The prompts on Hinge profiles, the limits on daily likes, and the emphasis on comments over swipes all serve a particular kind of user. That user is willing to pay.
Bumble’s decline suggests that a middle-ground approach, where the app tries to serve everyone, may be losing its appeal. Tinder’s slight revenue decline paired with heavy new spending tells a similar story from a different angle. Pouring money into product development makes sense when the underlying product still has demand. When users are leaving because they are tired, more features can feel like more of the same thing packaged differently.
Will 2026 Be Any Different?
Probably not in any dramatic way. The AI features rolling out will improve some parts of the process, particularly safety verification and spam reduction. Chemistry may help some users feel less overwhelmed by choice. But the core tension remains: people are fatigued by the format itself, and no algorithm resolves that on its own.
The apps that perform best in 2026 will likely be the ones that define their audience narrowly and build around a specific kind of relationship goal. Hinge already demonstrates this. Platforms trying to be everything to everyone will keep spending and keep losing ground, because the problem was never a lack of features. It was a lack of purpose.
