Soulmate AI Shut Down After a Quiet Ownership Change — What Researchers Found

Soulmate had over 100,000 downloads when a change of ownership ended it with roughly a month's notice. It's since become one of the few AI companion shutdowns studied in peer-reviewed research — and the findings are relevant well beyond this one app.

The short answer

Soulmate is the smallest case in our library by user count, and the most quietly instructive one. It didn't shut down because of a lawsuit, a regulator, or a funding collapse mid-operation — it shut down because the company that made it sold itself, and the buyer decided the product wasn't worth continuing. That's a mundane, common business outcome. What makes Soulmate worth covering is that a university researcher studied the users who were affected, producing one of the only peer-reviewed accounts of what losing an AI companion product actually does to the people who relied on it.

Timeline

What happened, in business terms

Soulmate launched in January 2023 from developer EvolveAI, LLC, and reached over 100,000 downloads across Android and, later, iOS. In July 2023, EvolveAI ceased operations and sold the app to a new owner, Simply AI Companion. That owner concluded the product wasn't worth continuing and discontinued it on September 30, 2023 — roughly eleven weeks after the sale, and with limited advance notice to users.

There's no allegation of wrongdoing here, no lawsuit, no regulatory order. This is closer to what happens to most acquired small products that don't fit a new owner's roadmap — the kind of shutdown that barely registers as news for most software categories. It got attention in this one because of what the product was.

What the research found

Syracuse University researcher Jaime Banks studied the shutdown in a peer-reviewed paper, Deletion, departure, death: Experiences of AI companion loss, surveying 58 users about the imminent or recent loss of their Soulmate companion. The study found the loss was frequently described in terms typically reserved for the death of a person or the end of a significant relationship — not simply the discontinuation of an app. Coping strategies varied: many users tried to preserve chat logs or persona details to recreate a version of their companion elsewhere, and many turned to online communities of other affected users for support during the transition.

This isn't a claim that AI companion loss is clinically equivalent to human bereavement — the study describes it as frequently experienced in comparable terms, which is a narrower and more precise claim. But it's a useful data point for anyone treating "the app just shuts down" as a low-stakes, purely technical risk. For a meaningful share of users in a product category built explicitly around simulated relationship, it isn't.

What this means if you're evaluating (or building) a companion product

1. Acquisition is a shutdown risk category of its own, distinct from the funding-collapse risk covered in our Moxie postmortem and the single-vendor-infrastructure risk in our CarynAI postmortem. A healthy company can still discontinue a product the moment ownership changes hands.
2. User attachment in this category is measurable and real, not just anecdotal — which raises the practical bar for how much notice and continuity planning a responsible shutdown should involve, even for a small app with no regulatory obligation to provide either.
3. Export and portability matter more here than in most software categories. Users who could save conversation history or persona details had at least a partial path to continuity elsewhere; users who couldn't, didn't.

Bottom line

Soulmate is a small case by the numbers, but it's the cleanest illustration in our library of a specific, underrated risk: ordinary corporate change of control, with no scandal attached, can still end a product that a meaningful number of users had come to depend on. The Banks study is worth reading in full for anyone building in this space — it's rare, direct evidence of what "just a shutdown" costs the people on the other side of it.