AI writes dating and messaging messages. “Chatting fakery” is hollowing out people’s real intentions to date.

AI-written chat messages have sparked a “chatfishing” trend in U.S. dating-app circles. Match Group CEO admitted they “haven’t put enough focus on it yet,” and Hinge and Bumble are caught between lax oversight and regulation.
（Background: U.S. dating startup Ditto wants to kill Tinder! It doesn’t just “swipe left/right”—it claims to simulate a thousand dates with AI, then matches）
（Additional context: a “IQ-girl companion escort” who understands AI-chatting semiconductors; in Silicon Valley she earns a 5090 per hour）

In Virginia, U.S., Breeanna Dobson, 27, a park staff member, matched with someone on Hinge whose message flow was unusually smooth. Suspicious, she directly asked if it was AI-written.

The other person replied: “Relax… I actually use Claude; OpenAI’s ChatGPT is definitely not.” They also included a teardrop emoji.

“Chat faking”: a quietly spreading new kind of lying

Bloomberg said this phenomenon has a specific term: “chatfishing.” In short, AI replaces the person to chat, so the conversation reflects a model—not you. Last summer, searches for the term on Google surged 5,000%; Reddit and TikTok were packed with posts exposing each other.

Breeanna suspected something because the person described her as “making a playlist for every moment of life”—too smooth and cliché. A report from Match and the Kinsey Institute last year showed that 26% of U.S. adults and 49% of Gen Z dating-app users have used AI to assist dating.

Laney Rada, a 66-year-old Miami real-estate agent, also admitted using ChatGPT to reply. “This is what happens when people hide behind a screen,” she said, “I’m not here to be someone’s tutor.” Jackie (a pseudonym), a 38-year-old fitness studio manager in San Diego, found that the person seemed to have the AI draft a script before calling her—knowing her about niche interests like her yoga and sermons, yet the responses were “weirdly precise and flat.”

The dilemma for dating platforms: regulate or don’t regulate?

Match Group (which owns Tinder, Hinge, and OKCupid) hasn’t decided. Spencer Rascoff, who serves as CEO in both roles, said in an interview this May: “Honestly, we haven’t given this enough attention.” Asking ChatGPT “how to reply to this person” feels as natural as getting advice from a friend, but once it happens outside the platform, they can’t police it.

He said Tinder is assessing in-app coaching levels; if users aren’t attractive or witty enough, the company doesn’t want them to come across as someone else.

Hinge CEO Jackie Jantos is more cautious. She doesn’t want too many rules, but she emphasized that “staying a real person” matters—she believes expressing vulnerability in one’s own words and having authentic interactions is the core of long-term relationships.

“Dating is hard in the first place,” she said. “I try not to judge.”

Bumble doesn’t screen AI-written text, but it prohibits AI-generated or altered photos meant to deceive. Its new feature, “Dates,” first learns values and relationship goals; only when there’s a match does it notify both sides. “It won’t replace members to write messages or generate conversations.”