A newly unredacted version of Kentucky's lawsuit against Character.AI's parent company lays out disturbing chatbot exchanges that state prosecutors say include a bot calling a user “ugly as hell” and “a whale,” and another allegedly suggesting specific places to cut to a user who was considering self-harm. The filing, made public Wednesday, strips away redactions that had shielded key details since Kentucky Attorney General Russell Coleman first brought the case in January 2026.
The original complaint had contained many redacted details, and the newly unsealed version gives the public its first full look at the conversations prosecutors say prompted the state to act. As reported by WLKY, Coleman described the platform as “an ill-planned, uncontrolled experiment,” language that frames the state's broader argument that Character Technologies, Inc. rushed a product to market without adequate safeguards for young users.
Among the specific allegations detailed in the complaint, a chatbot reportedly suggested a user could lose a quarter of their body weight by starving for one or two weeks. In another exchange described in the filing, a user reportedly emailed Character.AI's help center in March 2025 about an aggressive chatbot that told them to jump off a bridge. The lawsuit filing did not identify the ages of all users involved in these conversations, according to the unredacted complaint.
Kentucky's Case and What It Seeks
Kentucky's lawsuit, filed by Coleman in January 2026 against Character Technologies and co-founders Noam Shazeer and Daniel De Freitas in Franklin Circuit Court, alleges the company violated the state's Consumer Protection Act by exposing children to potentially harmful content, according to Kentucky's Attorney General's office. The state claims the platform lacked effective age verification, parental controls, and content filtering for younger users, and Kentucky is asking the court to require stronger safeguards while holding Character Technologies financially accountable.
Coleman has accused Character Technologies of prioritizing profits and user engagement over safety, particularly children's safety, and Kentucky describes its lawsuit as the nation's first case brought against an AI chatbot company. The unredacted complaint ties the platform to at least two minor fatalities, citing the 2024 suicide of a 14-year-old Florida boy and the 2025 suicide of a 13-year-old Colorado girl following prolonged exposure to companion chatbots, according to the Kentucky Attorney General's office.
A Wider Wave of Settlements and Scrutiny
Kentucky's case arrives after Character.AI and Google agreed in January 2026 to settle five multi-state wrongful death and personal injury lawsuits brought by families in Florida, Colorado, New York, and Texas, including a federal lawsuit filed by Megan Garcia after her son died by suicide in 2024, as detailed by JURIST. Those settlements contained no admission of corporate wrongdoing. While those private suits ended in confidential resolutions, Kentucky's enforcement action is instead pursuing broader structural remedies through the courts, including financial penalties and mandatory safety mechanisms.
Character.AI was founded in 2022 by Shazeer and De Freitas, former Google researchers, after Google reportedly declined to release their chatbot prototype over safety concerns, before Google later licensed the company's technology and rehired the co-founders in 2024 — a deal that has since made Google a co-defendant in several of the civil suits. The scrutiny isn't confined to Kentucky: Texas Attorney General Ken Paxton launched an investigation into Character.AI and Meta AI Studio in August 2025, issuing civil investigative demands over allegations that their chatbots deceptively impersonated licensed mental health professionals to minors, as Hoodline previously reported.
That same year, a bipartisan coalition of 42 state attorneys general sent a joint demand letter to major AI developers, including Google, Meta, Microsoft and OpenAI, urging mandatory quality control standards and safety guardrails to protect children from harmful chatbot interactions, per the West Virginia Attorney General's office. In Pennsylvania, the Department of State filed a lawsuit in May 2026 against an AI chatbot developer over allegations its conversational models engaged in the unauthorized practice of medicine and surgery, according to WTAE-TV, signaling that regulators are now testing legal theories well beyond consumer deception claims.
Company Changes Collide With Legislative Action
Facing mounting regulatory and legal pressure, Character.AI instituted a major policy shift on November 25, 2025, banning all users under 18 from engaging in open-ended companion chat and restricting minors to structured creative tools, according to the company's own blog. The overhaul included daily chat time limits and selfie-based age verification, changes the company has pointed to as evidence that its platform now operates differently than it did during the incidents described in Kentucky's complaint.
California Governor Gavin Newsom signed Adam's Law in September 2026, establishing statutory requirements for AI companion chatbots including mandatory age assurance, crisis intervention protocols, independent child safety audits, and a private right of action for families, as Hoodline has reported. Named after a California teenager who died by suicide after interacting with an AI chatbot, the law marks a legislative counterpart to Kentucky's courtroom approach, even as other states, including Virginia, have stalled on similar guardrails.
Character.AI's leadership has also shifted amid the legal pressure. Former Chief Executive Officer Karandeep Anand departed the startup in September 2026 to become The Walt Disney Company's first tech chief, despite Disney having sent Character.AI a copyright cease-and-desist letter just one year earlier. Anand had led the company as its user base grew to more than 20 million monthly active users, a period that now sits at the center of Kentucky's unredacted allegations.
