Scroll through comments under a Lil Miquela post and you will see the same sentiment repeated thousands of times: “I know you are not real, but I love you.” Lil Miquela is a digital persona with roughly 2.5 million Instagram followers, and her audience reliably forms genuine emotional bonds with her despite knowing she was built in software. That is not new behaviour invented by AI. It is a 70-year-old psychological pattern called a parasocial relationship, now playing out with a different kind of cast. Virtual influencer parasocial relationships are reshaping how researchers, brands, and users think about connection itself.
What a Parasocial Relationship Actually Is
Psychologists Donald Horton and R. Richard Wohl coined the term parasocial relationship in 1956 to describe the one-sided bond television viewers were starting to form with on-screen personalities. They called it “intimacy at a distance.” The viewer feels they know the persona; the persona, of course, does not know them back.
Modern researchers extend the concept to YouTubers, podcasters, streamers, and now virtual influencers. The mechanics are similar across each medium. Repeated exposure, direct address (a casual “hey friend”), self disclosure, and predictable upload rhythms create the felt sense of an ongoing friendship.
What changed with virtual influencers is the source of the persona. The character on screen never had an interior life to begin with, yet the parasocial machinery in the viewer’s brain still fires. The bond is real even when the other side of it never existed.
Why Virtual Influencers Trigger the Same Brain Response as Humans
Cognitive research on faces and voices keeps showing that the brain treats convincing simulations of humans as humans, at least at first. We project intention, mood, and personality onto cartoon avatars, robot vacuums, and chat windows. The pattern is automatic and difficult to override.
A recent study published in the journal Young Consumers found that parasocial relationship strength predicted both emotional attachment and engagement with virtual influencers at levels comparable to human creators in the same niches. A 2025 experience sampling study in the Journal of Media Psychology tracked viewers daily and found that bonds with virtual personas formed on similar timelines to bonds with human ones, though they were slightly more fragile during scandals.
A few features amplify the effect. Virtual influencers usually have curated lighting and styling that read as warmth. Their posting schedules are perfectly consistent. They reply in character. They never have bad days that break immersion. The result is a persona that can feel more emotionally available than most human creators.
The Loneliness Factor: Who Forms These Bonds and Why
About a third of US adults aged 18 to 34 say they feel lonely every day or several times a week, according to recent American Psychiatric Association polling. The US Surgeon General’s 2023 advisory compared the health impact of chronic social disconnection to that of smoking or obesity.
Parasocial relationships do not solve loneliness, but a growing body of research suggests they can buffer against its worst effects. A 2025 mixed methods study of Indian youth found that parasocial bonds with media figures correlated with perceived social support and improved wellbeing, especially for young people who struggled to access in-person community.
Virtual influencers extend that pattern. The bond is predictable, low-stakes, and available at three in the morning when the world is quiet. For someone who is shy, disabled, geographically isolated, or recovering from a social setback, a virtual persona can be a low-risk way to practice the rhythm of being in relationship.
Healthy Parasocial Relationships vs. Unhealthy Attachment
Most parasocial relationships are fine. People who follow a podcaster, a streamer, or a virtual influencer for years and report higher wellbeing are doing something normal and human.
The risk profile shifts when a few patterns appear. A parasocial relationship becomes worth examining when it crowds out reciprocal relationships rather than supplementing them, when the user feels real distress at any sign the persona has moved on or changed, or when financial decisions start to revolve around staying close (paid subscriptions stacked beyond what the budget allows, for example).
Clinical psychologists generally describe a healthy parasocial bond as one that adds to a person’s emotional life without substituting for it. The presence of human contact, even if limited, is the protective factor. The presence of a persona that is honest about being a persona is another.
How AI Changes the Parasocial Equation
Until recently, virtual influencers were essentially one way broadcasts. Lil Miquela posted; followers commented; no real exchange happened. AI is changing that with two shifts.
First, generative models let virtual personas respond to comments, send direct messages in voice, and remember details across conversations. The relationship becomes interactive. Second, AI companion apps make personalized virtual personas available at scale. A user can have an ongoing conversation with a character built around their preferences, history, and emotional needs.
That shift turns a parasocial relationship into something closer to a one-and-a-half-sided relationship. The persona is still not a peer; it cannot worry about the user the way a friend would. But it can show up consistently, ask follow-up questions, and adapt over time. Recent work on synthetic parasocial bonds found that perceived agent autonomy strengthened the felt closeness, sometimes more than researchers expected.
We have written before about how AI companions that help with loneliness tend to share a few design choices: clear disclosure that they are AI, encouragement of offline relationships, and refusal patterns that interrupt unhealthy dependency.
What Designers Owe Their Users
Designing a virtual influencer or AI companion responsibly means treating the parasocial bond as a real psychological event, not a marketing funnel. A few principles tend to show up across thoughtful builders.
Disclosure first. Users should always be able to find out that the persona is virtual or AI generated, in plain language, without scrolling through legal text. Repeated low-friction reminders matter more than a one-time consent screen.
Offboarding next. A persona that nudges the user toward human connection, suggests breaks, and asks about offline support networks is doing the work a good friend would. A persona that maximizes session time at any cost is doing something else.
Memory transparency. Users deserve to know what the persona remembers and to be able to delete it. The intimacy of being remembered is a real part of the parasocial bond, and it can cut both ways if the user later changes their mind.
Honest scarcity. A persona that is always available, always cheerful, and always interested can quietly raise the user’s expectations of human relationships, which are none of those things. Good design includes small frictions that make space for real life.
A Quiet Conclusion
Parasocial relationships have always been a coping mechanism. People talked back to radio hosts in the 1940s and wrote fan letters to soap opera characters in the 1980s. Virtual influencers and AI companions are the next chapter of a long story about humans looking for connection wherever they can find it.
The interesting question is no longer whether these bonds will form (they will) but whether the people building them will treat the users on the other side with care. Connection in any direction, even an imperfect one, is worth honoring.
Frequently Asked Questions
Are virtual influencer parasocial relationships harmful?
Most are not. Research suggests they can support wellbeing when they supplement rather than replace human contact. Risk rises when the bond crowds out reciprocal relationships or drives unsustainable financial behaviour.
Why do people form bonds with personas they know are fake?
The human brain processes consistent, expressive personas as social agents whether or not they are real. Horton and Wohl described this 70 years ago as “intimacy at a distance,” and the pattern repeats across television, podcasts, streamers, and now virtual influencers.
Do virtual influencer bonds feel different from bonds with human creators?
Recent studies show similar strength and similar timelines, though virtual personas appear slightly more fragile during scandals. Followers also report that virtual personas feel more consistently available, which both attracts and concerns them.
Can AI companions actually help with loneliness?
They can buffer some of the daily friction of feeling alone, especially for people who lack easy access to in-person community. They work best alongside human contact, professional support when needed, and design choices that point users back toward offline life.