Hinge, Tinder, Bumble: How A.I. Is Reshaping Dating Apps

Artificial intelligence is moving beyond chatbots and into matchmaking, changing how major dating platforms — and a new wave of start-ups — connect people. Reporting in The New York Times describes an emerging model in which A.I. matchmakers reduce endless swiping and deliver curated, paid matches instead.

The shift is illustrated by Emma Inge, a 25-year-old project manager in San Francisco who turned to a start-up called Known after years of swiping on apps such as Hinge and Tinder. Ms. Inge spent about 20 minutes with an A.I. matchmaker — essentially an A.I. chatbot — describing what she wanted and what she wanted to avoid. A week later she received a match and, for a one-time fee of $25, met him at a bar.

“With how dating is nowadays, I thought, ‘Oh, well, let’s try it,’” Ms. Inge said. “Let’s do it for the plot.” Her blind, A.I.-paired evening produced a two-hour conversation and exchanged numbers — but the human follow-up failed when she was later ghosted. “That was the kick of it, that the A.I. actually did find compatibility,” she said. “It was the human part that didn’t work out.”

Start-ups like Known — founded in May by Celeste Amadon and Asher Allen, who left Stanford to build the company — are experimenting with pay-per-date models and psychologist-designed questionnaires. Ms. Amadon says a pay-per-date approach is “more incentivizing to actually get people out in the real world and on dates,” and Known has run singles nights in San Francisco that drew several hundred attendees.

At the same time, the largest players in the industry are integrating A.I. into their products. Match Group, which owns Hinge and Tinder, is increasingly leaning on A.I.; Hesam Hosseini, Match Group’s chief operating officer, called the technology “the next technological shift.” Hinge uses A.I. to give profile feedback and has retooled its matching algorithm with generative A.I., which Mr. Hosseini said increased matches by about 15 percent.

Tinder is testing a system called Chemistry that would use A.I. to learn about users — including by scanning camera rolls if users opt in — and could be offered initially for free. Grindr has introduced six A.I. features it dubs gAI, including an A.I. “wingman,” tools to revive old matches and A.I.-generated profile summaries; Grindr’s chief executive, George Arison, plans to include popular features in a premium tier. Bumble has said it intends to release an A.I. matchmaking app by year’s end and is considering charging per match.

Executives and investors are watching closely. New leadership has accelerated some efforts: Match Group hired Spencer Rascoff last winter and he now oversees Tinder personally; Whitney Wolfe Herd, Bumble’s founder, returned as CEO in March and has said A.I. dating “is not a fad.” At the same time, private equity firms have approached dating apps about acquisitions as they seek ways to compete with Match Group’s scale.

The push to A.I. comes at a difficult moment for the industry. User satisfaction and the number of paying subscribers have declined: Bumble lost 9 percent of paid subscribers over the last year and Match Group lost 5 percent, even as overall user bases grew. Paid users remain a vital revenue source; Match Group’s paid 20 percent of users generate the majority of revenue. Shares for both companies are down significantly from earlier highs.

Companies are also grappling with user resistance to poorly executed automation and unwanted A.I. “slop.” Some firms downplay A.I. branding: Hinge, for example, avoids labeling certain features as A.I. internally. Other experimental tools under development include automated dating coaches and A.I. agents that could test features by interacting with each other, an idea already explored in reporting about A.I. clones.

Despite the risks, executives see upside in replacing the exhausting cycle of swiping — download, swipe, get ghosted, delete, repeat — with higher-quality, guided introductions reminiscent of early online dating questionnaires. Mr. Hosseini compared the potential shift to the era when sites like eHarmony used long questionnaires to match users.

The New York Times reporter Eli Tan documented these developments and the mixed early results: A.I. can identify compatible traits and surface promising matches, but human behavior remains unpredictable. As Hinge, Tinder, Bumble, Grindr and start-ups such as Known experiment with different models and monetization strategies, A.I. is poised to reshape what paying for dating looks like — even as the industry navigates regulatory, ethical and user-experience hurdles.