Reach Is Not a Business Model: What Character.AI's Bankruptcy and Disney's $20 Million Bid Reveal About AI Products

In mid-October 2026, Character Technologies, the company behind the chatbot platform Character.AI, filed for Chapter 11 bankruptcy. Court documents put assets and liabilities each at between $50 and $100 million. Disney is in as the first bidder: $20 million for a non-exclusive, perpetual license to parts of the technology and for certain legal claims, plus transition support from former employees Disney has already hired. The bid only sets a floor; other interested parties can offer more. The sale is expected to close in late November or early December; if no buyer emerges who will keep the platform running, it will be wound down. That's reported by TechCrunch and The Next Web, among others. The reports don't disclose revenue or monthly costs.

Users aren't the same as value

The contrast is striking. In 2022, users exchanged more than two billion messages with the bots in five months; by October 2024, the platform counted around 20 million monthly active users, and in August 2024 Google paid roughly $2.7 billion for a license to the company's models. Two years later, the same company is standing before a bankruptcy judge, and the first concrete bid for parts of it is $20 million. The company itself cites legal, regulatory, and operational headwinds. Add compute costs that rise with every use: a product that costs money to run every conversation needs revenue that grows with usage. Reach alone doesn't supply that.

The real subject: rights

What I find most instructive is what Disney wants to buy. It isn't the user base, and it isn't the platform as a whole, but a license and certain claims. The same company had told Character.AI in September 2025 to remove chatbots using Disney characters. A year later it's at the table, at the point where the conversation is about rights, not reach. That fits an observation I already made on the GEMA ruling against Suno: where generative systems work with protected content, what ultimately decides things is who can document a clean chain of rights.

What I take from this for studios and brands

First: anyone building a campaign or dialogue project on a platform should clarify what happens if that platform fails. Data that can be exported, content and characters in your own hands, contracts that survive a wind-down. It's the same question as with the retirement of an image model, only with higher stakes. Second: when doing brand and character work with AI, it pays to ask licensing questions before the first prototype, not after launch. Third, for us as a provider: an installation that delights people isn't yet a business. Whether it sustains itself depends on who pays, what each use costs, and who owns what gets created.

Where I stay cautious

The reports rely on court documents and trade media; figures on revenue and monthly losses are missing or only occasionally mentioned. How the proceedings end is open, as is whether other bidders will appear. And the reasons for the bankruptcy are multi-layered: lawsuits from families, youth-protection requirements, and age verification play into it just as much as the business model. Drawing a simple lesson about "AI bubbles" from this would be too convenient.

Conclusion

Character.AI shows that a large, engaged user base doesn't carry an AI product if costs, rights, and regulation don't grow along with it. Disney's bid says more about that than any user figure: what's valuable is what can be licensed and legally documented. For everyone building brand or experience projects with AI, it's a clear reminder to check a project's rights foundation and economics earlier than its reach.