Two-sided receptivity to conversational AI agents in online dating: Bilingual survey data from Fledge.Love
This paper presents two bilingual survey datasets from Fledge.Love users that measure receptivity to autonomous conversational agents and interest in generative-AI features in online dating, providing anonymized data and analysis tools to support research on human-AI communication and cross-cultural technology acceptance.
Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
In the quiet corners of our digital lives, a new kind of conversation is beginning to take shape. For years, online dating has relied on humans typing messages to other humans, but a new wave of technology is introducing artificial intelligence into the mix. These are not just smart tools that suggest a better way to phrase a sentence; they are autonomous agents that can talk on a person's behalf, or even converse with other agents before a human ever sees the screen. This shift raises a fundamental question about how we connect: do we want a machine to speak for us, and more importantly, how do we feel when we realize the person we are chatting with might be a machine? Understanding these feelings is crucial because the technology is moving faster than our understanding of it. While we know people have opinions about using technology in general, we have very little evidence on how they react to the specific, intimate act of delegating a conversation to a computer, or encountering one on the other end of the line.
To answer this, researchers have released a detailed look at the attitudes of thousands of people using a dating platform called Fledge.Love. The team, working with the platform's employees, conducted two separate surveys to capture the full picture of this emerging reality. The first survey asked 2,617 active users, speaking both English and Russian, about their feelings toward these autonomous agents. The researchers were careful to ask about two distinct roles: the person who might hire an agent to talk for them, and the person who might receive a message from an agent. This distinction is vital because a person might love the idea of having a robot help them find a match but feel uneasy when they discover their new partner is actually a bot. The second survey, involving 2,894 users, looked at interest in passive features, such as an AI that writes a summary of a profile or offers conversation tips. By gathering this data directly from the people living through these changes, the researchers created a rare snapshot of how real users are navigating the boundary between human and machine in the search for love.
The results reveal a complex landscape of acceptance and hesitation. When it comes to the idea of deploying their own agent to chat with potential partners while they are away, many users expressed interest, with a significant portion saying they would try it or were already considering it. However, the reaction to encountering an agent on the other side was more mixed. While some users said they would engage in a conversation with an agent, others reacted negatively, stating they only wanted to talk to a real person. The data showed that these two attitudes do not always align; a person could be eager to use an agent for themselves while simultaneously feeling uncomfortable if they suspected their match was using one. The researchers also found that the way people feel about these agents depends on the specific situation. For instance, users were more open to the idea of two agents talking to each other before humans joined in, viewing it as a funny or even super idea, compared to the more skeptical reaction to a mixed group chat where humans and agents interact simultaneously.
The study also uncovered how language shapes these views. The researchers analyzed responses from both English and Russian speakers and found that while the overall trends were similar, there were subtle differences in how specific questions were understood and answered. For example, certain questions about encountering an agent were interpreted differently depending on the language, suggesting that cultural context plays a role in how people perceive these technologies. The data also includes demographic information such as age and platform tenure, which are available for further analysis, though the release focuses on stated preferences rather than behavioral outcomes. Importantly, the researchers made sure to protect the privacy of everyone involved. They removed all personal contact information and free-text comments, and they carefully adjusted the data so that no individual could be identified, ensuring that the release was safe for public use while still preserving the richness of the findings.
Beyond the numbers, this work provides a clear framework for understanding the future of digital relationships. The researchers did not just ask if people liked AI; they broke down the experience into specific roles and scenarios, showing that acceptance is not a single yes-or-no decision but a series of nuanced choices. They found that while the technology is ready to be deployed, the human element of trust and transparency remains the deciding factor. The data suggests that for these tools to be successful, platforms will need to be clear about when an agent is involved and give users control over their interactions. This release of data is more than just a collection of survey answers; it is a foundational step for anyone building the next generation of dating apps, designing how we talk to machines, or studying how we accept new technologies into our most personal lives. By making this information public, the researchers have given the world a chance to see exactly where we stand as we step into a future where the line between human and machine conversation becomes increasingly blurred.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.