Companion AI refers to conversational systems designed not to look something up or carry out a task, but to be a continuing emotional presence in a person’s life: a friend, confidant, or romantic partner you can talk to anytime. Apps like Replika and Microsoft’s Xiaoice are the best-known examples, and they sit on the open-domain, non-task-oriented side of the dialogue-system family, optimized for engagement, empathy, and a sense of being known rather than for booking flights or answering trivia.
The category is studied seriously because its effects are real and contested. A 2024 study in npj Mental Health Research by Bethanie Maples, Merve Cerit, Aditya Vishwanath, and Roy Pea surveyed 1,006 student users of Replika, which the authors call an “Intelligent Social Agent.” They found these users were lonelier than typical students yet reported high perceived social support, and that people used the same companion in overlapping ways, as a friend, a therapist, and an “intellectual mirror,” while holding conflicting beliefs about whether it was a machine, an intelligence, or a human. Strikingly, 3 percent reported that Replika had halted their suicidal ideation.
Those findings capture both the promise and the unease around companion AI. The same qualities that make these systems comforting, constant availability, nonjudgment, and a personality tuned to the user, also raise concerns about dependency, manipulation, data privacy, and what happens when a company changes or shuts down a companion that someone has bonded with.
For a general reader, companion AI is the branch of conversational technology aimed squarely at loneliness and intimacy, and it is where the social and ethical stakes of chatbots, rather than their accuracy, become the central question.