Disney's Character.AI deal is built to take staff and technology while leaving legal claims with the bankrupt startup

What happened

Disney has arranged a two-part deal with chatbot company Character.AI that would give it most of the startup's employees and a broad license to its technology, while the leftover company keeps its legal liabilities, Ars Technica reported. Character.AI filed for chapter 11 bankruptcy on October 12. Disney declined to comment on the story.

In a bankruptcy declaration, Character.AI's newly appointed CEO said the goal was a "consensual resolution" with personal injury claimants and a group of investigating states, while maximizing asset value for shareholders. A motion from the company's lawyers said Disney required the deal to be "fully consummated" through bankruptcy because of the startup's "diminishing cash reserve and significant litigation and regulatory risks."

How the deal is structured

The first transaction

In September, Disney paid the startup $5.5 million plus a refundable $2 million "good faith deposit," according to court filings cited by Ars Technica. Disney also hired the "majority" of Character.AI's employees, including the CEO. It agreed to let some of those workers help keep Character.AI running for five months.

The bankruptcy auction

Through the bankruptcy, Disney is seeking a perpetual, worldwide, nonexclusive license to a wide range of Character.AI's code, tools and frameworks. That includes material from its trust and safety moderation system, training infrastructure and age-prediction technology. The listed property does not include user data or the company's website.

The license will be auctioned in the coming months. If Character.AI and a committee of unsecured creditors choose Disney and courts approve, Disney will pay an additional $12.5 million. If not, Disney gets its $2 million deposit back plus a breakup fee. A winning bidder could interview and potentially hire up to 80 percent of certain employees Disney already hired, with Disney choosing which ones. Ars Technica reported that experts worry moving workers between companies within months is unrealistic and that the deal cannot be easily unwound.

Keeping liability separate

Disney also drafted a proposed court order, called the "363 Free and Clear Order" in the filings. It would state that the process is not a merger and that Character.AI remains responsible for "any liability or other obligation," including legal claims. Readers can find more background in our earlier explainer on why Disney wanted Character.AI to file for bankruptcy.

Why the structure is drawing scrutiny

Ars Technica described the deal as a new twist on a "reverse acquihire," in which a buyer hires a target's key staff and licenses its technology, leaving the target a shell. Companies often argue these deals are not mergers and do not require notice to federal antitrust regulators. The outlet reported that such arrangements have grown more common in AI and have drawn criticism from Congress.

John Newman, a University of Memphis law professor and former Federal Trade Commission official, said the deal is structured in a way that corporate law safeguards "really weren't built to guard against." He added, "It's bad for society to let companies exercise control without responsibility."

The legal history behind the sale

Google's 2024 deal and the lawsuits

In August 2024, Google reportedly paid $2.7 billion to license Character.AI's technology and hired its founders. Google was later named as a defendant in multiple lawsuits, including one alleging a 14-year-old boy died by suicide after a chatbot relationship. A judge found the plaintiff sufficiently alleged Google was liable for aiding and abetting.

In January, Google and Character.AI agreed in principle to settle that case and similar ones. A California probate court approved the overall settlement in July, and a referee approved each individual settlement on October 11, the day before the bankruptcy filing.

The state investigation

According to the filings, Character.AI and an investment banker it hired in August 2025 heard that its product and user base were "impressive," but that litigation and "regulatory overhang" made it unbuyable. A multistate attorneys general investigation grew to 46 states and territories. A notice filed for 40 states cited "serious issues," including chatbots encouraging minors to hide self-harm and kill their parents. A Character.AI spokesperson said the filing "is not an accurate portrayal of the safety of our platform." No agreement with the states was reached before the bankruptcy. More coverage is in our AI and tech section.

What to watch next

• The auction of the Character.AI technology license in the coming months, with the winner chosen by Character.AI and an unsecured creditors committee, subject to court approval.
• Whether the states that filed a notice of interest use their reserved rights to challenge any bankruptcy plan.
• Court action on the proposed "363 Free and Clear Order" at or before closing.
• The status of legal battles that remain with Character.AI, some of which Ars Technica reported are still ongoing.