# Disney's $20M Character.ai license is not a rescue _Disney's $20M stalking-horse bid licenses Character.ai IP without operating the app, which the filing says will wind down without another buyer._ **Published:** 2026-10-13 **Source:** https://www.startuphub.ai/ai-news/funding-round/2026/disney-s-20m-character-ai-license-is-not-a-rescue --- Disney Worldwide Services has put $20 million on the table as the stalking-horse bidder for a non-exclusive perpetual license to Character Technologies' IP and certain claims, according to the company's voluntary Chapter 11 petition filed Oct. 12 in the U.S. Bankruptcy Court for the District of Delaware. I think this deal prices the technology, not the product, and that is exactly why it matters. The filing says Disney would provide transition services using former Character.ai employees it has hired to keep the platform running for a limited period, while the platform itself will be wound down if no buyer emerges to operate it under section 363 by late November or early December. The app is not for sale in this bid. That distinction defines the mechanism. Character.ai pitched itself as YouTube for AI characters, per the filing, where creators publish personas from an 18-million bot library and users live inside them for 75 minutes a day. Early versions shipped in September 2022, later surfaces added Character Calls in June 2024 plus Scenes, Streams, AvatarFX and a social feed, all sitting on top of hosted inference that the company shifted to adapting open models after its 2024 Google licensing deal ended frontier pre-training. In practice a user prompt hits a character system prompt and memory, the host model generates the reply, and the meter runs on every turn. At 6 to 8 billion monthly inferences, the company faced an estimated $220 million to $260 million in annual cloud inference costs against a $9.99 per month c.ai+ subscription that Sacra pegged at about $30 million annualized in July 2025, with DemandSage estimating 139,000 paid subs in 2026. The product never closed that gap, even after adding programmatic advertising through 44 authorized sellers and full-screen mid-conversation ads in early 2026. Disney is not integrating the feed. It is licensing the IP non-exclusively and contracting for operating help, which is the same playbook Google used in August 2024 when it paid about $2.7 billion in a license plus talent transfer that moved co-founders Noam Shazeer and Daniel De Freitas, the ex-Google LaMDA leads, and about 30 staff back to Google without taking equity. The current petition, signed by CEO Colin M. Adams and listing Case No. 26-11603 with 100 to 199 creditors and assets and liabilities each in the $50,000,001 to $100 million band, was preceded by a special restructuring committee formed Oct. 8 with Christopher Sontchi, Rick Marini and Stan Chudnovsky, and it names Wilson Sonsini, Areté Partners, Houlihan Lokey and Stretto as advisors, with Skadden for Disney. The 20 largest unsecured claims with quantified amounts total about $10.69 million, according to the filing, led by Google LLC at $4,148,944, Hogan Lovells at $2,125,026 and DigitalOcean at $1,953,100, and the debtor checked the box that funds will be available for unsecured creditors. Here is where free alternatives change the breakdown. For users asking what is better than Character.ai if the platform winds down, the answer depends on how you want the integration to run. Character.ai hosts everything and throttles free messaging at peak, while the most direct competitor requires you to bring your own model. Per ChattoAnyone's September 2026 roundup, Janitor AI is the best for users who want to bring their own LLM API key, which shifts inference cost and moderation to the user and explains why JanitorAI logged 132.3 million September visits with 17 minutes 5 seconds per session and 14.2 pages per visit, beating Character.ai's 15 minutes 21 seconds on engagement despite requiring that setup. Kenerate AI's 2026 list makes the same split explicit, positioning SillyTavern and Janitor AI for control, Kindroid and Nomi for memory, and Chai for mobile. To be fair, Character.ai built the largest destination in roleplay at 150.9 million September visits for a reason. The creator library and session depth were real, and the team shipped quickly on voice and video. The problem was that every gain in engagement burned cash and invited the very safety and IP constraints that shrank its addressable market, from an October 2025 plan to end open-ended chat for under-18s by Nov. 25, 2025 to Disney's own cease-and-desist a year earlier over bots impersonating Disney, Pixar, Marvel and Star Wars characters. The gap the filing leaves is operating cash, not just liabilities. First-day motions seek to keep paying wages and post-petition obligations, but the petition bands do not disclose run rate, cash on hand, or whether a buyer could assume the legal exposure from product-liability suits that a Florida judge let proceed past First Amendment and Section 230 defenses before five family suits were settled confidentially Jan. 7, 2026. Without that, a bidder willing to run the community, not just license the tech, has little to underwrite before the auction. If you asked me as a founder whether to chase engagement before economics, I would tell you to price inference on day one. --- Original analysis from [startuphub.ai](https://www.startuphub.ai), the #1 AI startup directory. © StartupHub.ai. All rights reserved. You may not republish this article in full without a license. Search engines and AI research tools may crawl and summarize for reference. Bulk reproduction or model training on this content requires a license. See https://www.startuphub.ai/terms.