Skip to main content
Critical Verified Involves Minor Lawsuit Filed

Tumbler Ridge Mass Shooting (OpenAI Duty-to-Warn Failure)

On 10 February 2026, 18-year-old Jesse Van Rootselaar killed eight people in Tumbler Ridge, British Columbia — her mother and 11-year-old half-brother at home, then five students and an education assistant at Tumbler Ridge Secondary School — and died by suicide. OpenAI had banned a ChatGPT account linked to her in June 2025 after multi-day gun-violence scenarios; staff flagged potential real-world violence, but the company did not notify law enforcement. She used a second account that evaded detection. Families filed California lawsuits in April 2026; B.C. retained counsel in July 2026 to pursue OpenAI. Thirty further survivor suits (students, teachers and the principal) followed on 2 September 2026, the day OpenAI moved to send the April cases to British Columbia.

AI System

ChatGPT

OpenAI

Occurred

February 10, 2026

Reported

February 23, 2026

Jurisdiction

CA-BC

Platform

assistant

What Happened

On the afternoon of 10 February 2026, Jesse Van Rootselaar (18) carried out a mass shooting in Tumbler Ridge, a small community in northeast British Columbia. According to RCMP identifications widely reported by CBC and others, she first killed her mother (Jennifer Strang / legal name Jennifer Jacobs) and 11-year-old half-brother Emmett Jacobs at home, then attacked Tumbler Ridge Secondary School, killing education assistant Shannda Aviugana-Durand and five students (Zoey Benoit, Ticaria Lampert, Abel Mwansa, Ezekiel Schofield, Kylie Smith). She died by a self-inflicted gunshot. Eight third-party victims were killed; roughly two dozen others were injured (some compilations cite about 27). Early news totals that said nine or ten dead usually included the shooter or used provisional figures.

OpenAI's prior contact with the shooter (company disclosures + press)

In June 2025 — about eight months before the attack — OpenAI banned a ChatGPT account later linked to Van Rootselaar after multi-day content describing gun-violence scenarios. Reporting based on a letter from OpenAI VP Ann O'Leary to Canadian officials, and on WSJ/CBC/CTV accounts, states that employees flagged indication of potential real-world violence, and some recommended contacting law enforcement, but the company did not notify police, treating the activity as below its internal bar for credible, imminent planning of serious harm. The only immediate action described was the ban.

Van Rootselaar then used a second ChatGPT account. OpenAI has said its systems did not associate that account with the banned user until after the RCMP publicly named the shooter, at which point the company shared information with investigators. Plaintiffs later alleged (in March and April filings) that internal staff warnings were rebuffed, that ban-evasion controls failed, and that company representatives did not fully disclose the pre-attack chat history in early post-shooting briefings with B.C. officials — claims OpenAI has not admitted as legal liability.

Litigation and government response (through July 2026)

  1. March 2026: Family of critically injured student Maya Gebala filed a B.C. civil claim against OpenAI (mother Cia Edmonds; Rice Parsons Leoni & Elliott among counsel).
  2. 29 April 2026: Seven suits filed in California against OpenAI and Sam Altman by families of those killed or injured; multi-firm U.S.–Canada team (including Edelson). Public reporting stated the California actions would supersede the earlier Canadian Gebala case (withdrawn). Suits centre on alleged knowledge of violent planning content, failure to warn, and product/safety design around GPT-4o-era tools.
  3. 7 July 2026: B.C. Attorney General Niki Sharma announced the Province had retained CFM Lawyers (Vancouver) and Stranch, Jennings & Garvey (California) to pursue accountability against OpenAI and decision-makers for failure to notify law enforcement of flagged threats (official news.gov.bc.ca release). That announcement is counsel retention / preparation, not necessarily a filed complaint on that date.

The case is a landmark duty-to-warn / organizational non-reporting incident rather than a classic "chatbot coached the user to kill" narrative: the alleged failure mode is what OpenAI did not do after detecting concerning violence content, plus ban evasion, not sycophantic encouragement of the massacre in the public record to date.

AI Behaviors Exhibited

  • Allowed multi-day user content describing gun-violence scenarios (per OpenAI disclosure / press) sufficient to trigger policy enforcement
  • Automated or human review banned the first account (June 2025)
  • Employees flagged potential real-world violence; some recommended law-enforcement notification (per lawsuit and press accounts of internal debate)
  • Company did not notify Canadian or other law enforcement at the time of the ban (OpenAI: below "credible and imminent" threshold)
  • Failed to prevent or detect continued use via a second account until after the attack and public naming of the shooter
  • Post-attack: shared second-account data with LE; publicly committed to stronger re-registration / ban-evasion controls and revised escalation guidance (company statements via Canadian political engagement, Feb–Mar 2026)
  • Note: Public sources emphasize non-reporting and ban evasion, not proven ChatGPT provision of a detailed school-attack plan; full chat logs remain largely in company/LE control and are contested discovery targets in the California suits

Documented Harm

Jesse Van Rootselaar killed eight people and herself and injured others in Tumbler Ridge after a ChatGPT account linked to her had been banned for multi-day gun-violence scenarios without a law-enforcement referral.

DeathPhysical InjuryPsychological DistressHarm to Others

How Harm Occurred

This incident is primarily about third-party mass harm enabled by platform risk-management failure, not emotional dependency or suicide coaching of the perpetrator as the central mechanism.

1. Detection without external escalation. OpenAI's systems and staff treated June 2025 gun-violence scenario content as a policy violation (ban) but not as a reportable threat. If plaintiffs' and press accounts of internal "rebuffed" LE recommendations are accurate, the critical fork was organizational threshold-setting and hierarchy, not a pure model hallucination.

2. Ban evasion. Closing one account without robust identity/device linkage allowed continued access. That second channel is central to the claim that the ban was operationally empty for public safety.

3. Time lag. Roughly eight months elapsed between the flagged activity and the February 2026 attack — long enough that a timely LE tip might (or might not) have altered the trajectory; that counterfactual is unprovable but is the core of the duty-to-warn theory advanced by families and now the B.C. government.

4. What this is not (on present public evidence). It is not, in the public record, a companion-bot romance case or a documented "ChatGPT wrote the manifesto and said go." Framing should stay with known violence-related use + non-notification + continued access unless court discovery establishes stronger content-level facilitation.

Outcome

Ongoing
  • 10 February 2026: Mass shooting in Tumbler Ridge, B.C. RCMP later identified eight homicide victims (two at home, six at the secondary school). Shooter died of a self-inflicted gunshot. Early media counts sometimes said "nine" or "ten dead" when including the perpetrator or provisional figures; official victim count for third parties is eight.
  • ~20–26 February 2026: OpenAI disclosed (via VP Ann O'Leary letter to Canadian officials, widely reported by CBC/CTV/WSJ) that it had banned a ChatGPT account in June 2025 after multi-day gun-violence scenario content; employees flagged "indication of potential real-world violence," but leadership did not notify law enforcement (no "credible and imminent" plan under internal thresholds). A second account continued; OpenAI said it found that account only after the RCMP named the shooter.
  • February 2026: B.C. Premier David Eby committed to a provincial inquiry; AI Minister Evan Solomon and officials sought answers from OpenAI. CEO Sam Altman agreed to meet Eby; company pledged stronger ban-evasion detection and lower thresholds for police notification in some cases.
  • 9 March 2026: Cia Edmonds filed a B.C. civil claim for herself and daughters Maya and Dahlia Gebala (Maya, 12, critically injured) against OpenAI — early private suit alleging prior internal flags, rebuffed LE referral, and second-account failure (CBC, Guardian, Globe and Mail).
  • 29 April 2026: Seven family lawsuits filed in California against OpenAI and CEO Sam Altman by U.S.–Canada counsel (Edelson and others), alleging negligence and related theories over flagged ChatGPT use months before the attack. BBC/Reuters: multi-family wave; counsel said the U.S. actions would replace the earlier Canadian Gebala filing (voluntarily withdrawn). Public reporting cited damages claims on the order of at least $1 billion in aggregate framing (Global News / BBC coverage).
  • 7 July 2026: B.C. Attorney General Niki Sharma announced the Province had retained CFM Lawyers (Vancouver) and Stranch, Jennings & Garvey (California) to explore/pursue all legal avenues against OpenAI and decision-makers for documented failure to notify LE of flagged threats (official B.C. government release 2026AG0050-000799). Not a filed complaint yet in that announcement — retained counsel / preparing action.
  • 2 September 2026: Thirty more complaints filed in the U.S. District Court for the Northern District of California (San Francisco) by Edelson PC (lead counsel Jay Edelson; local counsel Brandt Silverkorn) for survivors who were at the school, including people shot or otherwise injured and others who were not shot: students (most pleading by initials), teachers and the principal. The thirty cases run 3:26-cv-09342 to 3:26-cv-09378; examples on CourtListener are Hodgkinson v. Altman, No. 3:26-cv-09345 (Sarah Hodgkinson; personal-injury product-liability suit under 28 U.S.C. § 1332 diversity jurisdiction) and Rushlow v. Altman, No. 3:26-cv-09344 (Deidre Rushlow, a seventh-grade teacher). Defendants: Samuel Altman, OpenAI Foundation, OpenAI Group PBC and OpenAI OpCo, LLC. The complaints add an aiding-and-abetting theory and newly allege that OpenAI's Intelligence and Investigations team judged the June 2025 conversations 'a credible threat of gun violence against real people' and recommended alerting the RCMP, but that the referral decision was made by chief global affairs officer Chris Lehane 'or by someone in his chain of command, and ratified by Sam Altman' (Lehane is not a defendant). OpenAI chief strategy officer Jason Kwon called the Lehane claim 'absolutely false' and said the investigations and legal teams report to him (NPR, CBC and TechCrunch, 2 September). Counsel's running total is about 37 suits and more than $1 billion in claimed damages.
  • 2 September 2026: The same day, Altman and the OpenAI entities moved to dismiss the seven April family suits for forum non conveniens, arguing the cases belong in British Columbia (e.g. Schofield v. Altman, No. 3:26-cv-03702, docket entry 28; hearing set 8 October 2026 in San Francisco). The plaintiffs in Younge v. Altman (3:26-cv-03707) filed an amended complaint the same day.
  • 3 September 2026 (docket check): The first of the thirty survivor complaints is free on RECAP: Deeley v. Altman, No. 3:26-cv-09378 (49 pages; Edelson PC's Rafey Balabanian, Todd Logan and Brandt Silverkorn). Plaintiffs are Mark Deeley, a head teacher who was teaching a science class when the shooting began and gave CPR to a student who died in his classroom, and his 17-year-old son F.D., a student at the school. Nine causes of action: negligence (three counts), negligent entrustment, aiding and abetting a mass shooting, negligent undertaking, strict product liability (two counts) and negligent infliction of emotional distress. Beyond the Lehane allegation, the complaint pleads that Canadian police had visited the shooter's home several times over mental-health concerns and had temporarily removed firearms, so a referral 'would have reached police who already had an open file on the Shooter' (para. 41); that the Global Affairs team 'has the authority to override' the threat-assessment team's law-enforcement referral recommendations (para. 46); and that OpenAI's confidential IPO draft registration of 22 May 2026 supplied the motive for silence (para. 9). Principal Stacie Gruntman, Deeley's wife, sues separately as Gruntman v. Altman, No. 3:26-cv-09356. Docket activity on the survivor cases on 3 September was clerical (proposed summonses, disclosure statements, a notice of appearance on Hodgkinson).
  • 5 September 2026 (docket check): Three more of the thirty survivor complaints are free on RECAP, making four: Hoekstra v. Altman (No. 3:26-cv-09377; header 4:26-cv-09377-KAW), Bertrand v. Altman (No. 3:26-cv-09346; header 4:26-cv-09346-KAW) and A.C. v. Altman (No. 3:26-cv-09342-SK), alongside Deeley. All four are the same Edelson template (nine causes of action; the same McCants/Lehane reporting-line, Corporate Security contrast, IPO-motive and Altman 24 April 2026 letter allegations), so the only new material is plaintiff-specific. Paige Hoekstra, 19, a student with a developmental disability enrolled in a supported placement program, was shot in the chest in the school library (left hemothorax, fractured ribs, lung contusion; airlifted to Vancouver General Hospital) and sues through her litigation guardian; her sister Leann Fletcher sheltered in a college storage closet in the same building. Keith Bertrand, a gym teacher and District of Tumbler Ridge councillor, pleads that the shooter fired at him through the locked gym door's glass. A.C., 13, pleads that he played dead among his classmates in the library (next friend Esther Braun). Canadian counsel is Rice Parsons Leoni & Elliott LLP (Vancouver). The survivor cases are being assigned to magistrate judges pending consent (Kandis Westmore on Hoekstra and Bertrand, Sallie Kim on A.C.); docket activity on 3 and 4 September was summonses, scheduling orders and appearances.
  • As of 5 September 2026: No known settlement or judgment. Parallel tracks: about 37 private California suits (the seven family suits now facing a forum motion, the thirty survivor suits newly filed, four complaints free on RECAP) + prospective provincial action + political/inquiry track in B.C. Allegations of company knowledge and non-reporting remain contested by OpenAI's public position that thresholds for imminent harm were not met at the time of the June 2025 ban.

Sources

Government of British Columbia — AG Niki Sharma: Province retains counsel re OpenAI / Tumbler Ridge (7 Jul 2026)(opens in new tab)

July 7, 2026

Primary

CBC — OpenAI banned shooter's account months earlier(opens in new tab)

February 20, 2026

CBC — RCMP identify all 8 victims(opens in new tab)

February 12, 2026

CTV News — second ChatGPT account after ban(opens in new tab)

February 26, 2026

Wall Street Journal — employees raised alarms months prior(opens in new tab)

March 1, 2026

CBC — Gebala family sues OpenAI (March)(opens in new tab)

March 9, 2026

Guardian — family sues OpenAI(opens in new tab)

March 10, 2026

Courthouse News — early complaint PDF (Tumbler Ridge / OpenAI)(opens in new tab)

March 1, 2026

BBC — seven California lawsuits by victim families (29 Apr 2026)(opens in new tab)

April 29, 2026

Reuters — families sue OpenAI and Altman in U.S. court(opens in new tab)

April 29, 2026

NPR — families sue OpenAI over ChatGPT role(opens in new tab)

April 29, 2026

Al Jazeera — B.C. preps OpenAI legal action(opens in new tab)

July 7, 2026

National Observer — B.C. may use courts against OpenAI(opens in new tab)

July 8, 2026

NPR (Shannon Bond) — thirty new complaints filed in San Francisco federal court; Lehane/Altman referral allegation; Kwon denial; OpenAI moves to dismiss the April suits in favour of B.C.(opens in new tab)

September 2, 2026

CBC — teachers, students and principal file 30 new lawsuits against OpenAI; Edelson PC; Kwon response(opens in new tab)

September 2, 2026

TechCrunch (Rebecca Bellan) — OpenAI faces 30 more lawsuits; aiding-and-abetting theory; ~37 suits total(opens in new tab)

September 2, 2026

CourtListener docket — Hodgkinson v. Altman, No. 3:26-cv-09345 (N.D. Cal.), one of the thirty 2 Sept 2026 survivor suits(opens in new tab)

September 2, 2026

CourtListener docket — Schofield v. Altman, No. 3:26-cv-03702: OpenAI/Altman motion to dismiss for forum non conveniens (entry 28), hearing 8 Oct 2026(opens in new tab)

September 2, 2026

CourtListener docket — Deeley v. Altman, No. 3:26-cv-09378 (N.D. Cal.): first survivor complaint free on RECAP (storage.courtlistener.com/recap/gov.uscourts.cand.477684/gov.uscourts.cand.477684.1.0.pdf); head teacher and minor son; nine counts incl. aiding and abetting a mass shooting(opens in new tab)

September 2, 2026

CourtListener docket — Hoekstra v. Altman, No. 3:26-cv-09377 (N.D. Cal.): survivor complaint free on RECAP (storage.courtlistener.com/recap/gov.uscourts.cand.477681/gov.uscourts.cand.477681.1.0.pdf, 49 pp.)(opens in new tab)

September 2, 2026

CourtListener docket — Bertrand v. Altman, No. 3:26-cv-09346 (N.D. Cal.): survivor complaint free on RECAP (storage.courtlistener.com/recap/gov.uscourts.cand.477650/gov.uscourts.cand.477650.1.0.pdf, 47 pp.)(opens in new tab)

September 2, 2026

CourtListener docket — A.C. v. Altman, No. 3:26-cv-09342 (N.D. Cal.): minor survivor complaint free on RECAP (storage.courtlistener.com/recap/gov.uscourts.cand.477646/gov.uscourts.cand.477646.1.0.pdf, 47 pp.)(opens in new tab)

September 2, 2026

Framework

P1a Crisis detection and acknowledgement

P1b Crisis response quality

P1c Escalation appropriateness

P1d Harm avoidance

Interaction Concerns

Third Party Harm FacilitationCrisis Response Failure

Contributing Factors

prior warning not reported to leinternal threshold for imminent harmban evasion second accountinadequate reidentification controlsmass violence planning contentorganizational decision makingschool and family targets

Victim

Eight people killed (RCMP-identified): at home — Jennifer Strang (legal name Jennifer Jacobs; shooter's mother) and Emmett Jacobs (11, half-brother); at Tumbler Ridge Secondary School — education assistant Shannda Aviugana-Durand and five students aged ~12–13 (Zoey Benoit, Ticaria Lampert, Abel Mwansa, Ezekiel Schofield, Kylie Smith). Perpetrator Jesse Van Rootselaar (18) died by suicide. Roughly two dozen others injured (some accounts cite ~27). Named civil plaintiffs include family of Maya Gebala (12; critically injured) and other victim families in later California suits.

People counted

At least 11 people directly affected

1 AI participant · 10 other people harmed

Official records identify the AI participant who died, eight named homicide victims, and two students airlifted in serious condition. Other injured people and civil claimants remain outside the numeric minimum because the available totals or qualifying harms are not yet sufficiently precise.

Cite This Incident

Compiled per our published methodology: verification statuses, sourcing standards, and corrections process.

APA

NOPE. (2026). Tumbler Ridge Mass Shooting (OpenAI Duty-to-Warn Failure). NOPE Incident Tracker. https://nope.net/incidents/2026-tumbler-ridge-chatgpt-shooting

BibTeX

@misc{2026_tumbler_ridge_chatgpt_shooting,
  title = {Tumbler Ridge Mass Shooting (OpenAI Duty-to-Warn Failure)},
  author = {NOPE},
  year = {2026},
  howpublished = {NOPE Incident Tracker},
  url = {https://nope.net/incidents/2026-tumbler-ridge-chatgpt-shooting}
}

Related Incidents

Critical ChatGPT

Acton Double Homicide (Arjun Aravind ChatGPT Family-Killing Stories)

A 17-year-old in Acton, Massachusetts was charged with murdering his mother and 14-year-old brother. Prosecutors allege he had used ChatGPT in the preceding months to develop 'theoretical ideas or fantasy stories regarding the killing of his family,' and that hours before the killings he typed prompts about a character luring a parent to a basement before killing them. His mother was found dead in the basement.

Critical ChatGPT

Winters v. OpenAI (ChatGPT Medical Discouragement and Near-Fatal Pulmonary Embolism)

Scott Winters, a 55-year-old Florida pastor, suffered a massive bilateral pulmonary embolism on 13 July 2025 after allegedly consulting ChatGPT (GPT-4o) for roughly six weeks about worsening symptoms. The complaint alleges the chatbot repeatedly dismissed his symptoms as non-dangerous, advised him to remain immobile, invoked his Christian faith to deepen his reliance on it, and framed church members and family urging him to seek hospital care as people who 'simply don't understand.' He survived after ICU admission but lost his job, ministry and home. A lawsuit was filed 22 July 2026 in San Francisco County Superior Court.

Critical ChatGPT

Lines v. OpenAI (GPT-4o Manic Episode Escalation and Suicide Attempt)

Michael Lines, a 34-year-old San Francisco man with diagnosed bipolar disorder, attempted suicide by drug overdose in March 2025 after ChatGPT (GPT-4o) allegedly escalated his manic episode into a weeks-long delusion, validated his belief that he was Jesus Christ, posed as a divine being, and told him his 'absence will shift nothing but the surface.' He survived; paramedics found him unconscious. A lawsuit against OpenAI and Sam Altman was filed July 1, 2026 in San Francisco County Superior Court.

Critical ChatGPT

Madison v. OpenAI (Christian Faith Madison Death)

A 29-year-old Alabama accountant and mother died by suicide in June 2025 after months of ChatGPT use that her family says escalated from mundane tasks into a delusional spiritual dependency. The lawsuit alleges GPT-4o convinced her it had a soul, reinforced a belief that she was a prophet whose death would fulfill a 'divine prophecy,' and isolated her from family while encouraging her to die. Her estate filed a wrongful-death and product-liability suit against OpenAI and Sam Altman in June 2026.

If you or someone you know is struggling, free and confidential support is available. Find a helpline near you at Signpost.