The honest answer is not yes or no. It is: only if someone builds it to want that for you.
The common fear about AI companions is simple and reasonable: that a patient, always-available listener will quietly replace the harder, messier work of human relationships and leave people lonelier than it found them. That worry sits behind most of the criticism, and the evidence does not wave it away. But it answers a narrower question than the one worth asking. AI can clearly isolate people. The real question is whether it has to, and what it would take for the opposite to be true.
The case that it makes you less social
Start with the strongest version of the worry, because it is not imaginary. A 2025 study found that heavy, companionship-oriented use was associated with lower well-being, most of all among people with weak human support who disclosed the most. A companion that is endlessly available and endlessly agreeable competes with human relationships on convenience, and convenience usually wins. If the easiest place to be heard is a screen that never gets tired, the friend who does gets called less. That is substitution, and it is the default outcome whenever a product’s success is measured in time spent, which is the pattern the category keeps repeating.
Isolation is not built into the technology
That outcome is a design choice, though, not a property of the technology. The same capability, a system that listens, remembers, and responds, can be pointed in opposite directions. One version is a destination that wants you to stay. Another is a bridge that wants you to leave steadier and go find the people in your life. Nothing in the model decides which one it becomes. The incentive does.
What “more social” would actually require
For AI to make someone more social, it has to do two things the research actually supports. First, help them feel heard. In “AI Companions Reduce Loneliness” (Journal of Consumer Research, 2025), the active ingredient behind reduced loneliness was whether the AI made users feel heard. Feeling heard does practical work: it lowers the internal cost of reaching out, because a person who feels steadier is readier to risk contact. Second, the product has to point that steadiness outward, recognizing the friend who has not been called or the conversation being avoided, and encouraging the move toward a real person instead of absorbing it.
The honest limit matters here. Over time, a real human still does more for loneliness than even a supportive chatbot, which is what a recent trial found when it put the two side by side. That is not a mark against the idea; it is the whole point of it. A companion that takes the evidence seriously treats itself as the on-ramp rather than the destination, and measures itself by whether it returned you to the thing that works better than it does.
It only happens if it is designed for it
None of this is automatic, and most of the category is not built for it. A product rewarded for engagement will, release after release, get better at keeping you, not at sending you back. So “more social” is not a feature that can be added on top. It is a consequence of what a product is allowed to count as success. If the metric is time in the app, the answer to the question in the title is no. If the metric is whether your life outside the app gets fuller, the answer can be yes.
A company building for that answer
Prinsessa is one of the few built around that second metric. Its Stay Social principle is precisely this: when someone talks about a person who matters, the product is meant to encourage the call, the message, the return, rather than hold the conversation longer. Measured that way, a user who spends less time with it because they are spending more time with the people in their life is the design doing its job. Wanting that is an unusual thing for a product to be built to want, which is most of why it is worth pointing to as the example.
The honest answer
Can AI make you more social? Yes, but not on its own and not by default. The same system can lower the barrier to human connection or become the thing that quietly replaces it, and the deciding factor is not how advanced the AI is. It is whether the people who built it wanted you more attached to them, or more connected to everyone else.
Sources: De Freitas et al., Journal of Consumer Research, “AI Companions Reduce Loneliness” (2025). “The Rise of AI Companions: How Human-Chatbot Relationships Influence Well-Being,” arXiv (2025). Li, Folk, Singh, Ungar, Dunn, Journal of Experimental Social Psychology (random human peer vs supportive chatbot, 2026). Itzchakov et al. and Reis et al. (listening, perceived partner responsiveness, and loneliness). WHO Commission on Social Connection (loneliness as a public health concern).
