Dating App Users Draw a Line on AI
AI was a major topic at the GDI Conference in London. Speakers discussed where the technology can improve online dating and where it risks undermining trust. The main divide was between using AI to protect and match users, and allowing it to take part in their conversations.
Nathan Studholme, Head of Product at Feeld, and Alexandra Kilpatrick, Head of UX Research, presented findings from a survey of 9,000 people about the use of AI in dating. While the industry is keen to adopt AI, 63% of Feeld users viewed its use in dating apps negatively, rising to 84% among female users. Respondents were more comfortable with AI being used to detect fake profiles, scams, risky behaviour and harassment than to write messages or steer conversations. Some 43% of Feeld users wanted no AI at all, while 30% preferred AI features to be optional. Their main point was that AI should protect the dating experience, not participate in it.
Karima Ben Abdelmalek, CEO of happn, made a similar argument. She said we are at a fork in the road. We can automate connection until AI begins replacing people, or keep AI in a supporting role with humans at the centre. AI can improve matching, detect fraud and identify fake accounts. However, using it to write messages or conduct conversations could further damage trust in dating apps. Karima said 41% of happn users felt uneasy about the effect of AI on human connection. Only one in ten said they would consider dating an AI. Her recommendation was simple: AI should protect users and help them find suitable people, but it should never pretend to be one.
New academic research provides another view of how daters may respond to these tools. The study, AI on Dating Apps: Links Between Usage Frequency, Self-Efficacy, Self-Esteem, and Willingness to Use AI, was written by Alice Binder, Anja Stevic, Jaroslava Kaňková and Jörg Matthes.
The researchers surveyed 693 dating-app users aged between 16 and 27. They examined whether people who frequently use dating apps feel more able to understand and influence their algorithms. They also looked at whether that confidence affects their willingness to use AI. Frequent dating-app users were more open to using AI for writing messages, improving profiles and selecting potential matches. Users who felt more confident about controlling an app’s algorithm were also more willing to use AI for communication, curation, fraud detection and profile verification. Women were less willing than men to use AI for communication and profile curation.
The findings do not necessarily contradict the concerns raised at GDI. They suggest that familiarity with dating apps can make people more comfortable with AI.
