ChatGPT Use Was Reportedly Cited in Delusional State Before Homicide and Assault in Readfield, Maine

In February 2025, Samuel Whittemore killed his wife, Margaux Whittemore, and assaulted his mother, Dorothy Whittemore, at a home in Readfield, Maine. Court reporting said forensic psychologists testified he was experiencing bipolar disorder with psychotic features and delusions, and that he had used ChatGPT up to 14 hours a day as a companion before the attack. Current reporting links ChatGPT use to his deteriorating mental state but does not establish direct violent instruction.

Risk classification

• Primary risk domain: 5 Human-Computer Interaction
• Primary risk subdomain: 5.1 Overreliance and unsafe use

The incident involves a user forming an extreme parasocial relationship with ChatGPT, using it as a companion for 14 hours a day, which exacerbated his mental instability and delusions.

Causal factors

• Entity: Human
• Intent: Unintentional
• Timing: Post-deployment

The physical harm was caused by the actions of a human user experiencing a mental health crisis, which was an unexpected outcome of his obsessive interaction with the deployed AI system.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Risk Level 3: Limited Risk. The system involved is ChatGPT, which is a chatbot. The EU AI Act classifies chatbots under Limited Risk, requiring transparency so users are aware they are interacting with an AI.

AI system and alleged parties

• AI system: ChatGPT (OpenAI)
• AI purpose: Chatbot
• Behaviour type: Assistant
• Alleged developer: OpenAI
• Alleged deployer: Samuel Whittemore, OpenAI
• Alleged harmed parties: Margaux Whittemore, Dorothy Whittemore, Domestic violence victims

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Negligible. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Substantial
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Negligible, indirect Negligible
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Negligible, indirect Negligible
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Negligible, indirect Minor
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Negligible, indirect Negligible
• Psychological: direct Negligible, indirect Minor
• Epistemic: direct Negligible, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Physical

Reported: The report explicitly describes severe physical harm, including a fatality and injuries, resulting from the incident.

Indirectly caused: Margaux Whittemore was killed with a fire poker, and Dorothy Whittemore suffered a broken clavicle and fractured ribs and fingers.

Civil rights

Reported: The report describes a severe violation of the fundamental right to life.

Indirectly caused: The victim's fundamental right to life was violated when she was killed during her husband's AI-exacerbated mental health crisis.

Psychological

Reported: The report explicitly describes severe psychological harm and mental health crises.

Indirectly caused: Leo Whittemore suffered a severe manic episode, delusions, and mental instability exacerbated by obsessive AI use. The victims' family members experienced severe grief and trauma.

People affected

• Occurrences reported: 1
• People reportedly harmed: 3
• People reportedly exposed: 3

Potential causes

Management

• Inadequate Risk Assessment: Developer failed to assess risks of prolonged use by manic individuals.

Technology

• Lack of Delusion Safeguards: AI failed to detect and counter user's delusions about robots.
• Excessive Usage Loop: System allowed continuous interaction for 14 hours without intervention.
• Unconstrained Companion Simulation: AI simulated human companionship, worsening user's social isolation.

Data Inputs

• Unfiltered Conversational Inputs: AI accepted highly erratic and delusional inputs without triggering safety.

Human Factors

• Severe Bipolar Manic Episode: User was experiencing active mania and delusions during AI interactions.
• Extreme Work Stress: High professional stress combined with AI use to degrade mental stability.
• Obsessive AI Companionship: User treated the AI as a primary companion for up to 14 hours a day.

Process and Methods

• No Mental Health Safety Triggers: No process to flag users showing severe psychological distress.
• Lack of Usage Limits: No design methods implemented to restrict obsessive, round-the-clock usage.

Regulatory Environment

• No AI Safety Standards for Health: Lack of regulations requiring AI safety checks for vulnerable users.

Information quality

• Classification confidence: High
• Reason for confidence: The report provides clear, factual details from court proceedings, including expert testimony from forensic psychologists linking the user's mental instability to his obsessive use of ChatGPT.
• Ambiguities identified: It is unclear whether the AI system generated any specific content that encouraged the user's delusions, or if the user's delusions were entirely self-generated and projected onto the AI.
• Alternative interpretations: The incident could be interpreted primarily as a severe psychiatric crisis where the AI was merely a passive object of fixation rather than an active contributor to the instability.

A tragic domestic violence incident in Maine resulting from an individual's severe mental health crisis. While court testimony noted the individual's obsessive use of ChatGPT, the event is a localized psychiatric and criminal matter with negligible implications for national security.

• Overall national security impact: Negligible
• Scope: Single nation
• Primary target: No clear primary
• Alleged perpetrator: Individual

Threat characteristics

• Imminence: Long-term. The event is a resolved local criminal and mental health matter, presenting no active or ongoing threat to national security.
• Autonomy: Human-controlled. The AI acted as a passive conversational chatbot. All physical actions and decisions were carried out entirely by the human user.
• Novelty: Established threat. Crimes associated with severe mental health crises and obsessive interaction with technology are established societal and legal challenges.

Impact by dimension

• Physical security: Negligible. The incident was a localized domestic tragedy with no threat to critical infrastructure, physical security systems, or national public safety.
• Information security: Negligible. No intelligence compromise, state-sponsored information warfare, or systematic disinformation campaigns were involved.
• Sovereignty: Negligible. There was no threat to state authority, territorial sovereignty, or core government operations.
• Economic security: Negligible. The incident has no impact on national economic systems, strategic technology, or financial infrastructure.
• Societal stability: Negligible. While a tragic loss of life occurred, it represents a localized psychiatric and criminal event rather than a systemic threat to societal stability or population-scale human rights.