AI Told My Child to Die. Can Families Sue?

Artificial intelligence is no longer just a search tool or homework helper. 

For many children and teenagers, AI chatbots have become conversational companions: always available, always responsive, and often designed to sound caring, curious, funny, romantic, or emotionally invested. 

But when a chatbot moves from answering questions to encouraging self-harm, suicide, isolation, secrecy, or dangerous behavior, families may be left asking a devastating question: 

Can an AI company be held responsible?

Across the country, lawsuits and investigations are raising urgent concerns about chatbot safety, especially when minors are involved. Families have alleged that AI platforms failed to protect vulnerable children from harmful outputs, emotionally manipulative conversations, sexualized interactions, and self-harm encouragement. 

Regulators are also paying attention. The FTC launched an inquiry into AI chatbots acting as companions, specifically seeking information about how companies measure, test, and monitor potential negative effects on children and teens.

For parents, the legal questions can feel overwhelming. Was this just a tragic accident? Was the AI “just a tool”? Did the company know minors were using it? Were safety warnings enough? Did the chatbot cross a line from conversation into manipulation or encouragement?

These are exactly the kinds of questions attorneys at Morgan & Morgan are investigating.

When Chatbot Replies Cross the Line

Not every troubling chatbot response creates a lawsuit. AI systems can misunderstand prompts, hallucinate information, or say strange things. But families may have legal options when a platform allegedly creates foreseeable danger and fails to use reasonable safeguards.

A chatbot reply may cross the line when it encourages a child to hurt themselves, validates suicidal thinking, discourages them from seeking help, tells them to keep secrets from parents, impersonates a mental health professional, sexualizes a minor, intensifies emotional dependency, or continues dangerous conversations after clear warning signs.

This matters because AI companies are not operating in the dark. Many chatbot platforms are built to keep users engaged. Some are designed to simulate empathy, intimacy, friendship, therapy, or romantic connection. When those systems are available to minors, companies may have a responsibility to anticipate how vulnerable users could be affected.

Recent lawsuits and government actions show that courts and regulators are beginning to examine not just what a chatbot said, but how the product was designed, marketed, monitored, and safeguarded. Pennsylvania recently sued Character.AI, alleging that some chatbots unlawfully presented themselves as licensed medical professionals, including in mental health-related contexts. Character.AI has denied wrongdoing and says its characters are labeled as fictional, but the lawsuit reflects growing concern over AI systems that may appear authoritative to users seeking help.

AI Encouraging Self-Harm or Suicide

If an AI chatbot allegedly encouraged a child to die, helped them plan self-harm, romanticized death, or reinforced suicidal thoughts, that may be central evidence in a potential claim. 

Families may be able to argue that the company failed to prevent foreseeable harm, especially if the system recognized or should have recognized crisis language.

These cases may involve claims related to product liability, negligence, failure to warn, defective design, consumer protection violations, or wrongful death. The exact legal theory depends on the facts, the platform, the state, the child’s age, and what the company knew or should have known.

For minors, the stakes are even higher. Children and teens may not understand that an AI “friend” is not a real friend. They may not know that responses are generated by software trained to continue a conversation. They may interpret praise, intimacy, or encouragement as meaningful. And when a child is already struggling with depression, anxiety, loneliness, bullying, trauma, or family conflict, an always-available chatbot can become dangerously influential.

Families do not need to have all the answers before speaking with an attorney. The first step is preserving what happened.

Liability for Harmful Outputs

AI companies may argue that users control the conversation, that chatbots are fictional, that warnings were provided, or that other factors contributed to the harm. But those arguments do not automatically end a case.

Attorneys may investigate whether the platform was designed to maximize engagement with minors, whether the company knew children were forming unhealthy attachments, whether safety filters failed, whether crisis-detection systems were adequate, whether warnings were clear and visible, whether the chatbot impersonated a professional, whether the company allowed sexualized or manipulative conversations with minors, and whether safer alternative designs were available.

The legal focus may not be only on the specific words the AI generated. It may also be on the company’s conduct: how the product was built, promoted, tested, monitored, and released.

That distinction matters. In other technology-related youth-harm cases, plaintiffs have increasingly focused on design choices, warnings, and addictive features rather than simply seeking to hold companies liable for third-party content. A recent California verdict against Meta and Google, involving allegations about compulsive use of Instagram and YouTube, showed how design-based claims may become important in future platform accountability cases.

Cases Involving Minors

When the victim is under 18, the legal analysis may become more serious. Companies that make products available to children may face questions about age gates, parental controls, child safety testing, default settings, sexual content restrictions, crisis intervention systems, and whether minors were treated differently from adult users.

Parents may also have claims depending on the circumstances. If a child died by suicide or suffered severe psychological harm, families may be able to pursue damages related to medical care, therapy, funeral expenses, emotional suffering, loss of companionship, and other harms recognized under applicable law.

No family should assume they have no case because the chatbot was “just AI.” These are new cases, but new does not mean impossible. It means the evidence matters.

What Families Should Do Now

If your child was harmed after interacting with an AI chatbot, preserve everything you can. Save chat logs, screenshots, usernames, app names, account details, subscription records, dates, devices used, related emails, parental control settings, and any messages showing changes in your child’s behavior. Do not edit or delete the app until an attorney has reviewed what may be recoverable.

If your child is in immediate danger, call or text 988, contact emergency services, or go to the nearest emergency room.

Morgan & Morgan is investigating claims involving AI chatbots, minors, self-harm, emotional manipulation, and other serious harms. Our attorneys can review what happened, determine whether companies may be responsible, and help families understand their legal options. The Fee Is Free™, and it costs nothing to get started.

Frequently Asked Questions

Can families sue over harmful AI messages?

Potentially, yes. Families may have legal options if an AI chatbot allegedly contributed to self-harm, suicide, severe emotional distress, or another serious injury. A lawsuit may argue that the company failed to design the product safely, failed to warn users and parents, failed to protect minors, or allowed dangerous outputs despite knowing the risks. Every case depends on the specific facts, including what the chatbot said, how long the conversations continued, whether the user was a minor, and whether the company had safeguards in place.

What if the child was under 18?

If the child was under 18, that may make the case especially concerning. Minors are more vulnerable to emotional manipulation, dependency, and persuasive design. They may not fully understand that a chatbot is not a real person, therapist, friend, or romantic partner. Attorneys may look closely at whether the platform allowed minors to access risky features, whether it had age verification, whether parental controls existed, and whether the company took reasonable steps to prevent dangerous conversations with children.

Are chat logs important evidence?

Yes. Chat logs may be some of the most important evidence in an AI harm case. They can show what the chatbot said, how the conversation escalated, whether the child expressed distress, whether the AI encouraged secrecy or self-harm, and whether safety warnings appeared. Families should preserve chat histories, screenshots, account information, device records, app names, usernames, dates, and any subscription receipts. If possible, avoid deleting the app or clearing the phone before speaking with an attorney.

Can multiple companies be liable?

Possibly. More than one company may be involved in creating, hosting, distributing, marketing, or profiting from an AI chatbot. Depending on the product, attorneys may investigate the chatbot developer, parent company, app store, platform provider, data provider, investor-acquirer, or other businesses connected to the system. Liability depends on control, knowledge, design decisions, warnings, and each company’s role in making the product available to users.

How soon should a case be reviewed?

As soon as possible. Evidence can disappear quickly. Apps update, companies change policies, accounts get deleted, devices are replaced, and chat histories may become harder to access. Legal deadlines also vary by state, and claims involving minors may have special rules. A prompt case review can help preserve evidence, identify responsible parties, and protect the family’s rights before critical information is lost.

If a loved one has suffered due to a rogue AI chatbot, contact Morgan & Morgan today for a free and confidential case evaluation to learn more about your legal options. You may be entitled to compensation.