Bumble dropped its women-message-first rule. Now dating apps are betting on AI to fix what swiping broke

Key points

• Bumble, launched in 2014, abandoned its signature rule this week requiring women to send the first message after a match.
• Dating app downloads and paid subscriptions have fallen across the industry as users report exhaustion with endless swiping.
• Companies including Bumble and Match Group are introducing AI features to prompt conversations and suggest compatible matches.
• No peer-reviewed study has yet shown that AI matchmaking produces better or longer-lasting relationships than conventional swiping.
• The shift places AI at the centre of some of the most personal decisions people make, with little independent scrutiny of how these systems work.

Bumble built its name on one idea. When a man and a woman matched, the woman had to speak first. That rule, in place since the app launched in 2014, was designed to cut down on unwanted messages and hand women a measure of control. This week, Bumble dropped it.

The move signals something bigger than a policy tweak.

Why are dating apps changing so fast right now?

Users are burning out. The industry has a name for it: swipe fatigue. Years of flicking through profile photographs, matching with strangers, and watching conversations fizzle before a first date has left many people feeling worse, not better, about their prospects. Downloads are falling. Subscribers are cancelling.

Dating companies are now, as the Guardian reported, in what insiders call "salvage mode." Their answer, almost universally, is artificial intelligence, the set of technologies that lets computers find patterns in large amounts of data and make predictions or decisions.

What are the apps actually doing with AI?

The features vary, but the pitch is consistent: let AI do the exhausting work for you.

Some platforms are testing AI that drafts opening messages on a user's behalf. Others use it to rank and filter profiles before a person ever sees them, predicting compatibility from behaviour, stated preferences and past choices. A few are experimenting with AI-powered "date coaches," chatbots (software that holds a typed conversation) that offer advice on what to say next.

These are company announcements, not clinical trial results.

No independent research published in a peer-reviewed journal, the standard test for whether a medical or psychological claim actually holds up, has yet confirmed that any of these AI tools increase the chance of a meaningful relationship. The apps measure engagement: time spent, messages sent, subscriptions renewed. Whether users end up happier is a different question, and one the companies have not answered publicly.

Should users trust AI to help them date?

Scepticism is reasonable. A short, honest answer: possibly helpful for reducing tedious admin, unproven for finding genuine compatibility.

AI systems learn from past behaviour, which means they can quietly reinforce existing biases, favouring certain body types, ethnicities or social backgrounds, depending on what patterns the training data contained. Dating app algorithms are proprietary, meaning companies do not have to disclose how they work.

For anyone using these apps, a few practical points are worth keeping in mind. An AI-drafted opening message is still a message sent in your name, so it shapes first impressions you did not personally craft. Features marketed as free trials sometimes lead to paid tiers. And "AI-recommended" does not mean scientifically validated.

Bumble abandoning its founding rule may turn out to be a sensible refresh, or it may simply be a struggling platform chasing headlines. Either way, the AI features being rolled across the dating industry deserve the same scrutiny we would give any other health-adjacent technology making promises about human wellbeing.