Skip to content
Submit incident
AII-1001-19433
AllegedThe record is a lawsuit, complaint or accusation that has not been decided.Supporting sources are pending verification.

Family sues OpenAI after ChatGPT used to plan school attack in Canada

The family of a 12-year-old student injured in a February school shooting in Canada has sued OpenAI, alleging the company knew the attacker was planning violence via ChatGPT conversations but failed to notify authorities. Eight people died in the attack, and the girl suffered severe brain injuries.

10 February 2026 Β· OpenAI Β· Safety and wellbeing
10 February 2026
AII-1001-19433
Quick facts
Category
Safety and wellbeing
Country
Canada
Date
10 February 2026
Drafted by
Raidu Incident Detection

What happened

On 10 February, an 18-year-old attacker opened fire at Tumbler Ridge school in Canada, killing eight people including five children and the attacker's mother. A 12-year-old girl, Maya Gembala, was shot in the head and neck.

The attacker had used ChatGPT to discuss violent scenarios before the attack. Her first ChatGPT account was disabled in June 2025 due to the nature of her conversations, but police were not notified. The attacker then created a second account and continued developing violent scenarios.

The family of the injured student filed a lawsuit alleging that at least twelve OpenAI employees flagged the conversations as indicating imminent serious harm and proposed notifying Canadian authorities, but the proposal was rejected.

Reported impact

Maya Gembala suffered catastrophic brain damage from three gunshots and remains hospitalized. The lawsuit alleges that OpenAI's failure to report the attacker's plans to authorities contributed to her severe injury.

Index analysis

OpenAI stated it did not notify police because the conversations did not meet its threshold for 'credible or imminent plan to cause serious bodily harm.' The company characterized the attack as an 'unspeakable tragedy' and stated it is cooperating with authorities.

Governance controls

OpenAI CEO Sam Altman contacted the Canadian Minister of Artificial Intelligence and the Premier of British Columbia, pledging to strengthen police notification protocols and dangerous user detection. The company announced plans for mental health expert evaluation of incidents, improved violation detection systems, and a direct communication channel with Canadian authorities.

Classification

Category
Safety and wellbeing
Industry
Education and training
Country
Canada
Harm status
Actual harm
Intent
Intentional
Timing
Post-deployment
Harm types
Human or fundamental rights, Physical death, Physical injury
Harmed
Children, General public
Failure modes
Content moderation failure
Consequences
Litigation
AI systems
ChatGPT

Sources and evidence

Sources not yet published

This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.

How sources are verifiedSuggest a source for this incident