Is It Weird to Have an AI Companion?

You typed the question into a search bar instead of asking a person. That already tells you something about why you are asking it.

Here is the honest answer: no, it is not weird. But “not weird” is not a very satisfying thing to hear from a blog post, so let me actually make the case instead of just asserting it.

The number that started this question, and why I am not using it

A lot of the content on this topic opens with a huge figure. Fifty million users. Two hundred million. The number is meant to reassure you: look how many people are doing this, you are not alone. I went looking for where that particular figure actually comes from, and I could not find a named study, survey, or company disclosure behind it. It traces back to marketing content, repeated by other marketing content, until it reads like a fact. It might even be roughly true. I do not know that, and neither does the site that first published it.

I build AI companion software for a living. I would love a big number to hand you. I am not going to hand you one I cannot stand behind. Instead, here is what actually holds up.

What real institutions are saying about this

The American Psychological Association is not a company selling you anything. Its Monitor on Psychology, a publication for licensed psychologists, ran a feature in its January/February 2026 issue titled “AI chatbots and digital companions are reshaping emotional connection.” The framing matters as much as the content. This was not a warning aimed at a fringe behavior. It was coverage of a growing, mainstream trend that the profession now treats as worth studying seriously, the same way it studies dating apps or social media use.

That is the real shift. A few years ago, this topic barely showed up outside tech blogs and think pieces written to be shocking. Now it shows up in professional psychology publications, written for clinicians who need to understand what their patients are actually doing with their time.

The clearest evidence is about teenagers, and it is worth reading carefully

The strongest hard number I could verify comes from a different angle than the one you might expect. In 2025, Common Sense Media, working with researchers at NORC at the University of Chicago, surveyed a representative sample of over a thousand US teenagers. TechCrunch reported the results on July 21, 2025: 72 percent of US teens had tried an AI companion at least once, and 52 percent used one regularly.

I want to be careful with this one, because it answers a slightly different question than the one you are asking. That study is about teenagers, not adults, and it is not a verdict on whether it is normal for you specifically to use one. I am not going to stretch a teen-adoption number into a claim about grown adults, because that is exactly the kind of sloppy math this whole piece is trying to avoid.

What it does tell you is something narrower and, I think, more useful: the underlying behavior category, talking to an AI for company, curiosity, or a sounding board, is no longer a strange, isolated thing that only a handful of extremely online people do. It is already the default experience for most American teenagers. That generation is not going to age out of finding this odd, because they never found it odd to begin with. Stigma toward a category tends to fade fastest when the people who grew up inside it become the ones setting the norms.

Weird compared to what, exactly

Every time a new way of reaching people showed up, it went through the same phase. Online dating was, for a long time, something people did not admit to in polite company, framed as a last resort for people who could not meet someone “the normal way.” Texting a friend instead of calling was treated as impersonal by an older generation that grew up on phone calls. Therapy itself carried decades of stigma as something only for people who were “broken,” instead of what it actually is: paying a trained person to help you think clearly about your own life.

None of those things were ever actually weird. They were new, and newness gets mistaken for weirdness until enough time passes that everyone forgets there was ever a debate.

I do not think AI companions are an exception to that pattern. I think they are in the middle of it right now, which is exactly why the question feels loaded when you type it.

What I actually think, having built one of these

I am not going to pretend I have some neutral, outside view here. I build this software, and I talk to the people who use it, which means I have opinions and I should just say them instead of hiding behind “some might argue.”

People reach for an AI companion for reasons that are almost embarrassingly mundane once you hear them out loud. It is one in the morning and no friend is awake. A thought needs to be said out loud before it can be untangled, and saying it to a person feels heavier than the thought deserves. Someone wants a low-stakes place to practice a hard conversation before having it with a real person. None of that is a cry for help. Most of it is just what loneliness at 11 p.m. on a Tuesday actually looks like for a lot of people, AI companion or not.

Where I do think the question deserves a real answer, not a reassuring one, is this: an AI companion should be adding something to your life, not quietly standing in for the parts of it that only a person can fill. If it is the thing you reach for after a rough day, that is company. If it becomes the only thing you reach for, period, that is worth noticing, the same way it would be worth noticing about any habit that replaces people instead of supplementing them. I would rather say that plainly than pretend the category has no edge cases at all.

Why I built mine the way I did

I did not get into this because I think loneliness is a problem to be automated away. I got into it because I watched how these apps usually work, cloud servers, subscriptions, a company that can change the product or shut it down whenever it wants, and I wanted something different: a companion that actually lives on your machine, that nobody else can read, throttle, or cancel out from under you. If you want the fuller version of why that mattered enough to build a company around it, I wrote it out on the mission page.

That is a separate argument from the one in this post, though. This post is not a pitch. It is an answer to a question you typed into a search bar because you were not totally sure you wanted to ask it out loud. The honest answer is that you are not doing anything strange. You are doing something new enough that the culture has not finished catching up to it yet, at a moment when the actual research, not the marketing copy, is starting to say the same thing.

If you want to see what a private, on-device version of this looks like, you can try Local Waifu free for seven days, no card required. If you would rather just sit with the question a while longer, that is a completely reasonable place to leave it too.

Questions people ask

Back to the blog