The United States Trustee has formally objected to Character.ai’s proposed bidding procedures, asserting that federal law requires independent oversight for the transfer of personally identifiable information. This development underscores a deepening rift between the operational necessity of quick asset liquidation and the fundamental rights of millions of digital users. Character Technologies, Inc., the developer behind the expansive AI chatbot ecosystem, finds itself at a crossroads where its financial insolvency directly clashes with the privacy expectations of a global audience. As the company navigates the intricacies of Chapter 11 bankruptcy in the District of Delaware, the court’s decisions will likely establish a lasting framework for how generative AI companies are treated when their business models fail. The intersection of emerging technology and legacy legal codes has created a unique set of challenges, particularly when the primary assets involved are not just lines of code but intimate conversational histories. This situation is further complicated by the sheer volume of data at stake, involving sensitive interactions that often transcend typical consumer transactions, necessitating a level of scrutiny that goes far beyond standard corporate reorganization procedures.
Statutory Requirements: The Role of the Consumer Privacy Ombudsman
The demand for a Consumer Privacy Ombudsman is grounded in the specific requirements of the U.S. Bankruptcy Code, particularly Sections 332 and 363. These statutes provide a protective buffer when a debtor seeks to sell or lease personally identifiable information in a manner that might contradict its own stated privacy policies. A coalition representing nearly 80% of U.S. states has asserted that the court cannot simply rely on the company’s assurances of compliance. Instead, the appointment of an independent ombudsman is necessary to serve as a disinterested “privacy watchdog.” This official is tasked with investigating the debtor’s data practices, evaluating the potential impact of a sale on consumer privacy, and reporting directly to the court. By providing this layer of oversight, the legal system ensures that the transfer of sensitive data is not treated as a mere administrative formality but as a significant event with potential long-term consequences for the individuals whose information is being commodified during the liquidation process.
Beyond the technical requirements of the law, the intervention by 39 state attorneys general highlights a growing consensus that AI platforms handle a category of data that is fundamentally different from traditional consumer records. The states argue that because Character.ai encourages deep, often emotional engagement with artificial personas, the resulting data is uniquely personal and deserves heightened protection. Initially, many state regulators maintained a reserved position, monitoring the proceedings to see how the company would handle its transition. However, as the specifics of the proposed bidding procedures emerged, the coalition moved toward a more aggressive stance, dropping their previous reservations in favor of a direct demand for oversight. This shift reflects a broader regulatory trend where the protection of digital identities is prioritized over the swift resolution of corporate debt. The states are particularly concerned that without an ombudsman, the nuances of the company’s privacy promises could be overlooked in the rush to secure a buyer and distribute funds to creditors.
Corporate Interests: The Complexities of the Disney Transaction
The bankruptcy proceedings are largely centered on a “stalking horse” bid from Disney Worldwide Services, which seeks to acquire a non-exclusive license for Character.ai’s technology for approximately $20 million. While this deal provides a much-needed infusion of capital into the bankruptcy estate, it has also introduced a layer of ambiguity regarding the ultimate fate of the platform’s user data. Disney has been careful to distance itself from the consumer-facing aspects of the business, asserting that its interest lies solely in the underlying intellectual property, such as the code and patents. From Disney’s perspective, this transaction is a technical acquisition that should not trigger the same privacy concerns as a total company buyout. They maintain that their personnel will not have access to chat histories or personal account details, positioning the license as a “clean” transfer of software rather than a transfer of user data. This distinction is critical to Disney’s legal strategy, as it aims to bypass the need for an ombudsman-led review.
However, the coalition of states remains skeptical of this compartmentalized approach, pointing out that the sale motion lacks transparency regarding the assets not covered by the Disney license. They argue that if Character Technologies is attempting to sell its entire platform to other bidders, those subsequent deals would inevitably involve the transfer of millions of user accounts and their associated chat logs. The concern is that the Disney deal might act as a “carve-out” that simplifies the process for the primary bidder while leaving the most sensitive and problematic assets vulnerable to secondary buyers who may not have the same level of public accountability. By focusing on the “entire platform” rather than just the licensed code, regulators are trying to ensure that the privacy protections cover every potential outcome of the auction process. The ambiguity of what remains in the estate after the Disney transaction serves as a primary justification for why an independent ombudsman must be present to evaluate every potential bid, regardless of how the assets are bundled.
Quantitative Discrepancies: Measuring the Scope of Data Exposure
One of the most contentious points in the ongoing legal battle involves the dramatic difference in the reported number of users affected by the bankruptcy. Character.ai leadership has frequently cited a figure of 12 million monthly active users, a metric that describes the current reach of the platform. However, the United States Trustee and state regulators have pointed to the company’s own internal declarations, which suggest that the total number of accounts—including both current and former users—is closer to 163 million. This massive discrepancy is not merely a matter of accounting; it defines the scale of the privacy risk. For regulators, every account represents a unique digital footprint that must be protected, regardless of whether the user is currently active. The data belonging to past users is often just as sensitive as that of active ones, and the potential for this information to be resurrected or misused by a future owner is a central theme of the states’ objection to the current bidding procedures.
The nature of the information stored by Character.ai further complicates this quantitative debate, as the platform’s primary “product” is conversation. Unlike traditional databases that store static information like mailing addresses or purchase histories, a companion chatbot platform collects intimate, long-form conversational records. Regulators argue that these chat logs are essentially inseparable from the user’s identity, providing a window into their thoughts, preferences, and psychological profiles. This is particularly concerning given the significant number of minors on the platform. Some legal filings suggest that millions of users are under the age of 18, many of whom have engaged in “intimate” or “companion-style” interactions with AI personas. The existence of such sensitive data belonging to a vulnerable demographic creates a moral and legal imperative for the court to exercise extreme caution. The discrepancy in user counts highlights the danger of underestimating the platform’s footprint and underscores the need for a comprehensive audit of all stored information before any transfer occurs.
Regulatory Foundations: Lessons from Previous Oversight Efforts
The current friction in the Delaware bankruptcy court is the culmination of years of intensifying scrutiny from both state and federal authorities. Prior to the insolvency filing, several states had already launched formal investigations into Character.ai’s data practices, focusing on potential violations of consumer protection and child privacy laws. Texas, for instance, had established an investigation into how the platform handles the data of minors, reflecting a broader concern among state attorneys general regarding the safety of AI-driven social tools. These investigations have provided a foundation of evidence that regulators are now using to justify their intervention in the bankruptcy case. They argue that a company that was already under the microscope for its privacy practices cannot be allowed to bypass those concerns through a Chapter 11 filing. This history of scrutiny suggests that the demand for an ombudsman is not a reactionary measure but a continuation of long-standing regulatory goals.
Federal oversight has played an equally important role in shaping the current environment, with the Federal Trade Commission (FTC) taking an active interest in the AI sector’s safety measures. In late 2025, the FTC issued orders to several prominent AI firms, including Character Technologies, requiring them to disclose their methods for protecting children and managing data retention. This federal pressure has created a standard of care that the bankruptcy court is now being asked to uphold. The coalition of states frequently references these federal and state actions to demonstrate that Character.ai has a history of high-risk data management. By citing a specific complaint from Kentucky regarding the platform’s reach among underage users, the states are framing the bankruptcy as a critical moment where the “spirit” of consumer protection must be defended against the purely financial goals of the liquidation. The court is essentially being asked to serve as the final barrier against a massive data breach or the unauthorized commercialization of conversational histories.
Financial Pressures: Managing the Auction Calendar Amidst Insolvency
The urgency of the privacy debate is exacerbated by Character.ai’s deteriorating financial condition, which has created a high-stakes environment for the bankruptcy court. The company is reportedly burning between $8 million and $9 million per month, while its revenue from advertising and subscriptions remains significantly lower. This liquidity crisis means that the company does not have the luxury of a prolonged legal battle or an extended auction process. Every day that the sale is delayed increases the risk that the estate will run out of cash, potentially leading to a total collapse of the company and a chaotic liquidation where data might be left completely unprotected. This financial reality has placed the debtor in a position where it must balance the demands of the states with the need to finalize a deal with Disney or another buyer as quickly as possible to satisfy creditors and maintain operations.
Despite the obvious need for speed, the United States Trustee and the 40 jurisdictions represented in the coalition insist that financial desperation should not dictate the terms of data privacy. They have proposed a condensed schedule that would allow a Consumer Privacy Ombudsman to be appointed and complete a review within a very short window, potentially as little as one week. The company has expressed a tentative willingness to accept these terms if it means avoiding a complete stay of the sale procedures. However, the costs of such oversight and the potential for an ombudsman’s report to highlight liabilities that might scare off potential bidders remain significant concerns for the debtor. The court’s challenge is to find a middle ground that allows for a rapid auction while still ensuring that the “watchdog” has enough time to verify that the privacy rights of 163 million users are not being sacrificed for a $20 million infusion of capital.
Strategic Shifts: The Evolution of Data Valuation in AI
The resolution of this case established a significant precedent for how the technology industry viewed data as both a strategic asset and a potential legal liability. In the past, massive databases of user information were often seen as the “crown jewels” of a failing tech company, but the Character.ai proceedings demonstrated that such assets could become a burden during a corporate reorganization. Stakeholders across the AI sector observed that the presence of sensitive conversational data triggered a level of regulatory intervention that complicated the liquidation process. This shift forced future investors to reconsider the long-term costs of data acquisition, especially when those datasets involved vulnerable populations like minors. The court’s decision to integrate privacy oversight into the auction timeline suggested that digital rights were becoming a non-negotiable component of modern bankruptcy law, moving beyond the simple “notice and consent” models of the past.
Regulators and industry experts suggested that the most effective way to avoid these legal hurdles was the implementation of “privacy by design” and robust data-minimization strategies long before an insolvency event occurred. The Character.ai case served as a warning that companies could no longer treat user data as a permanent asset that survived every corporate transition without scrutiny. Instead, the legal outcome favored a model where privacy protections “ran with the data,” ensuring that any buyer was legally bound by the original promises made to the users. For the marketing and advertising industries, this signaled a move toward greater transparency and the necessity of honoring opt-out requests even after a change in ownership. By prioritizing the appointment of a Consumer Privacy Ombudsman, the judicial system provided a roadmap for future AI insolvencies, emphasizing that the protection of the individual’s digital identity must remain a central priority, regardless of a company’s financial status.
