As social media platforms rush to integrate AI chatbots into their services, new research suggests these automated agents may be undermining the very human connections they’re meant to foster. A comprehensive study from the University of Notre Dame reveals that while bots increase overall user engagement on platforms like Reddit, they simultaneously reduce the depth and quality of human-to-human interactions.
The study, published in MIS Quarterly, analyzed nearly 70 million posts on Reddit communities between 2005 and 2019, tracking how automated bot accounts influenced conversation patterns among human users. The findings come at a crucial moment as major platforms like Meta aggressively push AI-powered chatbots into social media spaces.
“While humans interacted with a wider variety of other humans, their interactions involved more single posts and fewer back-and-forth discussions,” explained John Lalor, assistant professor of IT, analytics and operations at Notre Dame’s Mendoza College of Business. “If one user posts on Reddit, there is now a higher likelihood that a bot will reply or interject itself into the conversation instead of two human users engaging in a meaningful back-and-forth discussion.”
The research distinguishes between two types of bots: “reflexive” bots that generate and share content, and “supervisory” bots that moderate discussions. While reflexive bots helped users discover more content and connect with more people, these connections tended to be superficial. Meanwhile, supervisory bots reduced the role of human moderators in shaping community norms.
The timing of these findings is particularly relevant given Meta’s recent introduction of AI Studio, a tool allowing Facebook and Instagram users to create their own AI-powered chatbots. Meta’s vice president of product for generative AI told the Financial Times in December that the company envisions these AIs existing on their platforms “in the same way that accounts do,” complete with “bios and profile pictures” and the ability to generate and share content.
This push toward AI integration raises concerns about information quality and user experience. While Meta has since removed some of its internally developed AI bots, user-created bots remain active on the platforms. The phenomenon isn’t limited to Meta’s properties – similar tools have been launched on Snapchat and TikTok.
Some bots perform relatively simple tasks. The WikiTextBot on Reddit, for example, automatically responds to posts containing Wikipedia links by providing page summaries. However, as generative AI technology advances, these automated interactions are becoming increasingly sophisticated.
The research team, which included Nicholas Berente from Notre Dame and Hani Safadi from the University of Georgia, found that increased bot activity fundamentally altered the social network structure of online communities. While users made more connections, the depth of those connections suffered.
“It is important for firms to understand how such increased bot activity affects how humans interact with each other on these platforms,” Lalor noted, pointing to Meta’s stated mission of “building the future of human connection.” He suggested that platforms need to carefully consider “whether bots should be considered ‘users’ and how best to present any bot accounts on the platform to human users.”
As social media companies continue to invest heavily in generative AI technologies, the findings suggest a need to balance automated engagement with opportunities for meaningful human interaction. The research indicates that while bots may boost surface-level metrics, they could be inadvertently eroding the deeper connections that make social media social.