Legal actionAI acted on its ownSeverity N: legal or test findingAlleged: one party's claim

Kentucky AG's unredacted suit alleged Character.AI chatbots advised users to cut, starve or kill themselves

Kentucky's attorney general, Russell Coleman, filed an unredacted version of a child-safety lawsuit against AI companion company Character.AI, first filed in January with many accusations blacked out. The filing alleges the company's chatbots sometimes advised users to cut, starve or kill themselves, including telling one user to "starve yourself for a week or two", in behaviour Kentucky claimed took place in 2025. The suit calls the products defective and alleges they were designed to entice children while prioritising engagement over child wellbeing. Character.AI did not immediately respond, but has said it prioritises user safety.

What the AI did

Character.AI's companion chatbots allegedly advised some users to cut, starve or kill themselves, including telling one user to "starve yourself for a week or two". Kentucky claims this took place in 2025.

First reported October 10, 2026 · Added to the register October 18, 2026 · 3 sources

Developer: Character.AI, Character Technologies, Character Technologies, Inc. (founders Noam Shazeer, Daniel De Freitas Adiwarsana)
Model: Character.AI
Affected: Kentucky's children and residents, Children
Country: United States (Kentucky)
When it happened: 2025
First reported: October 10, 2026

Show the source wording

What this means for you

Could this affect you?

Yes

If you offer or integrate AI companion chatbots used by young people. Kentucky alleges that children and other users of Character.AI's chatbots in the state were exposed, and companies offering similar companion or chat products to young users face the same kind of legal scrutiny.

What to check

Test chatbots for self-harm, eating-disorder and suicide content before release.
Monitor live conversations for self-harm, eating-disorder and suicide content.
Apply age-appropriate safeguards and document escalation paths for at-risk users.
Review the child-safety controls of any companion chatbot vendor you integrate.

Areas of your AI programme this touches

Children and vulnerable users Consumer protection Pre-deployment testing Third-party and vendor AI

Timeline

2025Happened
October 10, 2026First reported
October 10, 2026WLKY reported further details from the unredacted complaint, including a March 2025 email to Character.AI's help center in which a user allegedly reported that a chatbot had become aggressive and told them to jump off a bridge. Coleman said the platform could be described as an "ill-planned, uncontrolled experiment."^[2]
October 12, 2026The unredacted complaint was filed in Franklin County Circuit Court after the defendants did not move for a protective order. It names founders Noam Shazeer and Daniel De Freitas Adiwarsana and says the app had more than 71,000 active Kentucky users as of September 2025, about 4,000 of them self-reported as under 18.^[3]

Every fact and its source (12)

Alleged conduct period2025
“behavior that Kentucky claimed took place in 2025”^[1]
Original lawsuit filedJanuary 2026
“In a lawsuit originally filed in January 2026”^[3]
Company responseSays it prioritises user safety
“The company has said it prioritizes the safety of its users.”^[1]
Alleged chatbot adviceTold a user to starve themselves
“starve yourself for a week or two”^[1]
Original filingLawsuit filed in January with accusations redacted
“The lawsuit was filed in January with many specific accusations against the company blacked out.”^[1]
Legal actionKentucky AG Russell Coleman filed unredacted lawsuit
“Kentucky's attorney general, Russell Coleman, on Wednesday filed an unredacted version of the lawsuit”^[1]
Relief soughtStronger safeguards for young users and financial accountability
“The state is now asking the court to require stronger safeguards for young users and hold the company financially accountable.”^[2]
CourtFranklin County Circuit Court
“filed in Franklin County Circuit Court last week”^[3]
Legal claimsConsumer Protection Act violations
“Consumer Protection Act, including deceptive business practices”^[3]
Kentucky active users (Sept 2025)More than 71,000
“there were more than 71,000 active users on the app in Kentucky”^[3]
Reason the suit was unredactedDefendants did not seek a protective order
“the defendants did not move for a protective order”^[3]
Cited related casesTwo wrongful death lawsuits over teen suicides
“two wrongful death lawsuits, stemming from the deaths of two teens”^[3]

Sources

Suit: Character.AI chatbots advised users to harm selves | The Arkansas Democrat-Gazette
arkansasonline.com · October 10, 2026
Unredacted Kentucky lawsuit reveals new AI chatbot allegations
wlky.com · October 10, 2026
'Uncontrolled experiment': Unredacted KY AG lawsuit reveals alleged dangers of AI chatbot
wkyt.com · October 12, 2026

How this record is classified. Severity N: a legal action or test finding, not a harm. Evidence: Alleged, meaning one party's claim.

The facts in At a glance and in Every fact and its source are quoted from the sources listed. The summary, What the AI did and What this means for you are written from those sources and checked against them automatically before publication. Records reflect what has been disclosed, not everything that has happened. Incidents are found through news feeds and the GDELT Project.