Disney's Character.AI deal leaves lawsuits behind

Disney hired most of Character.AI's staff and bid $20 million for a license to its technology in bankruptcy, leaving lawsuits and user data behind.
What it means for founders
- Liability follows the company, not the code. Google's 2024 license and hiring deal, about $2.7 billion, did not keep it out of court; a judge let aiding and abetting claims proceed. Disney's deal cannot close without a court order, which Character.AI is seeking, cutting off successor claims. Buying a whole company, as AMD agreed to do with World Labs, means buying its history too.
- Unsettled safety claims destroy exit value. A product with millions of engaged users drew only a $20 million bid for a non-exclusive license, after the roughly 28 parties a banker approached cited litigation and regulatory overhang. For companion apps, settled cases and working safety controls now shape valuation.
- Chat logs are a regulated asset. A privacy policy that permits transfer in a sale, a gap Ars Technica flagged, will not settle it. Build deletion, export and account level opt-outs now.
- Watch November. Proposed dates put bids due November 20, an auction November 23 and a sale hearing November 24, with a December 17 outside date.
The story
Disney has taken on most of Character.AI's employees, among them former chief executive Karandeep Anand, who TechCrunch notes is now Disney's first CTO. It is also the lead bidder to license much of the chatbot maker's code. Court filings show lawsuits, state investigations and user data stay with the bankrupt company.
How the deal is built
Character Technologies filed for Chapter 11 in Delaware on October 12. Its sale motion lays out two linked stages. Under a September 18 agreement signed before the filing, Disney paid $7.5 million up front, including a $2 million refundable deposit, and hired most of the workforce. Some of those hires help run the platform under a transition services agreement for up to five months from that date. Next, Disney would pay the rest of a $20 million total for a permanent, global, non-exclusive license covering code such as the trust and safety moderation stack, age prediction and training infrastructure.
The motion says Disney required the deal to close inside Chapter 11, citing the startup's shrinking cash and its litigation and regulatory risks. Disney assumes no liabilities, the license would pass free of successor liability, and Character.AI releases Disney and the hired staff from claims. User data, chat logs, training data, finished models and the consumer app are excluded. If another bidder wins and closes, Disney gets its $2 million deposit back plus a break-up fee and expenses capped at $700,000 combined, subject to court approval.
Disney declined to comment to WIRED, which quoted former FTC official John Newman: "It's bad for society to let companies exercise control without responsibility."
What Character.AI still faces
The chief executive's first day declaration says earlier injury suits are settled, but one newer suit, cases brought by Kentucky and Pennsylvania, and a 46 state and territory investigation remain open.
On October 14, 40 attorneys general filed a notice of interest asking that any buyer adopt strong age assurance, crisis safeguards, limits on training models with chats and a chance for users to delete data before a sale. Character.AI's chief executive, in a statement carried by The Verge, said the filing misrepresents the platform's safety. Its blog says the service stays online but will be wound down if no bidder wants to run it.
What we don't know yet
- Whether any bidder will keep the Character.AI app running.
- Who might buy the user data and chat logs, and under what limits.
- Whether the states will object to the releases and free and clear terms.
Sources
Primary sources
Enki Daily
Get stories like this every weekday morning.
The day's AI stories for founders, each with what it means for your company. Free.
More in Funding & Business
- Tenstorrent raises $1.5B at a $6.2B valuation

For founders: Another inference supplier now has volume money. If you serve open-weight models such as DeepSeek or Kimi, ask your inference hosts whether they plan to offer…
TechCrunch · 21h ago - Anthropic pledges 5% of its shares to a benefit fund

For founders: There is no new grant program to apply to yet. The closest live money is the $200 million research fund, which takes proposals of at least $1 million, mostly…
TechCrunch · 1d ago - Asia's Q3 startup funding hits $38.5B, 55% to AI

For founders: Seed money is easier to find than later rounds suggest. Seed and angel totals more than doubled year on year, so a credible AI angle can open early…
Crunchbase News · 4d ago - Anthropic says it hires fewer juniors and meets more

For founders: Hiring plans may tilt senior. If this pattern spreads, entry-level engineering roles get scarcer and candidates who can run agents with little oversight get…
TechCrunch · 2d ago - Replit makes OpenAI default, keeps Claude as option

For founders: Pin the model you trust. If you build with Replit Agent and liked what Claude produced, open the selector in Power or Max mode and choose it explicitly.
TechCrunch · 3d ago