News & Analysis

OpenAI Faces School Shooting Lawsuits: AI Liability Questions

By David Meldofsky

Published May 2, 2026 · Updated May 19, 2026

Two separate sets of mass-shooting lawsuits now name OpenAI: the seven Tumbler Ridge federal suits filed in California in April 2026, and the Chabba family's eight-count complaint filed in the Northern District of Florida in May 2026. They arise from different shootings, rest on different evidentiary records, and plead partly different theories. Read side by side, they show as much about how AI-harm litigation is being structured as about how it is being argued.

This article compares the two case sets directly: what each alleges, how their evidence and posture differ, and what that divergence means for how each is likely to be litigated and resolved. The doctrinal questions these cases share with the rest of the OpenAI docket — how courts will characterize generative output, how Section 230 and the First Amendment apply, and whether any duty to warn exists — are covered in the framing section of the central hub, OpenAI Lawsuits. This page stays on the two shootings specifically.

What the lawsuits claim

The Tumbler Ridge lawsuits arise from the February 10, 2026 shooting in British Columbia. Reuters reported that seven federal lawsuits were filed in San Francisco by families affected by the shooting, with the Associated Press noting claims that include negligence, wrongful death, and product liability theories.

The complaints, citing earlier Wall Street Journal reporting, allege that OpenAI's internal safety team flagged the user's conversations and recommended notifying police months before the attack, but that recommendation was not acted on. They also allege a specific product-design failure: after one account was disabled, the user allegedly created another and continued using ChatGPT.

The account-circumvention allegation is the one doing the doctrinal work. A generalized "the company should have monitored its users better" theory is diffuse and invites an early dismissal; an allegation that the enforcement system itself could be defeated by simply re-registering is concrete and points at how the system was engineered and operated rather than at the content of any particular exchange. That is the version of the claim built to survive a motion to dismiss.

Plaintiffs' counsel Jay Edelson is among the more sophisticated tech plaintiffs' lawyers in the country and has signaled the lawsuits are part of a broader effort. That does not speak to the merits, but it suggests these cases are being structured for staying power, not quick settlement.

What's doctrinally at stake

Both case sets run into the same threshold dispute that every suit in this docket has to clear: whether a generative system's output is treated like a manufactured product, with traditional tort doctrine applying, or like hosted content, with the platform-immunity and speech defenses in play. That characterization tends to drive the Section 230, First Amendment, and duty-to-warn questions rather than those being resolved on their own terms.

Because that framing analysis is shared across the whole docket and is not specific to the shootings, it lives in one place rather than being repeated here: the framing section of the central hub, OpenAI Lawsuits, sets out the product-versus-content question, how Section 230 and the First Amendment map onto it, and where the duty-to-warn theory sits. What follows stays on what is genuinely different between the two shooting case sets.

A separate Florida case raises overlapping but distinct issues

The Tumbler Ridge lawsuits are not the only U.S. civil action of this kind. On May 10, 2026, the family of Tiru Chabba — one of two people killed in the April 17, 2025 mass shooting at Florida State University — filed an eight-count federal complaint in the Northern District of Florida against OpenAI and the individual alleged in the complaint to have carried out the shooting. The case is captioned Joshi v. OpenAI Foundation et al., Case No. 4:26-cv-00222-MW-MJF. Plaintiffs' counsel are Osborne Francis & Pettis in Florida and the Strom Law Firm in South Carolina, with civil rights attorney Bakari Sellers on the trial team. For a full breakdown of the complaint, the eight counts, and the chat-history allegations, see the dedicated article on the Chabba FSU shooting lawsuit.

For purposes of this comparison, the first real difference is the extra count. The Florida complaint adds negligent entrustment, which the Tumbler Ridge filings do not plead. Its appeal is structural: it reframes the dispute around whether OpenAI can be answerable for continuing to grant a user access once warning signs appeared, which is a question a court can reach without first deciding the harder one about how to classify the system's output. Whether that doctrine stretches to mass-market software access is unsettled, and the answer will carry past this case.

The second real difference is the evidence each set rests on. Tumbler Ridge, citing Wall Street Journal reporting, alleges OpenAI's internal safety team specifically flagged the user before the attack. The Florida complaint does not allege pre-shooting internal flagging. It relies instead on chat logs obtained by Florida law enforcement after the shooting, combined with a body of published reporting and OpenAI's own statements documenting system-wide safety lapses (New York Times reporting on Microsoft's pressure to ship faster, CNBC reporting on shortened safety-testing windows, Metr's public statement that it had insufficient evaluation access to the o3 and o4-mini models, and OpenAI's own admissions about the GPT-4o sycophancy rollback). Tumbler Ridge is the stronger specific-knowledge case; the Florida case is the stronger systemic-negligence case. The two evidentiary records will likely produce different settlement dynamics even though the underlying defendant is the same.

Florida Attorney General James Uthmeier had previously announced a criminal investigation into OpenAI and ChatGPT related to the same FSU shooting. Criminal investigations operate under different standards from civil claims and do not establish wrongdoing. The convergence — civil lawsuits in two U.S. jurisdictions, state criminal scrutiny, prior regulatory attention, and a separate California wrongful death suit filed the same week alleging ChatGPT gave fatal drug interaction advice (see the Turner-Scott overdose lawsuit) — signals AI safety questions have moved out of academic debate.

OpenAI has responded on the record. Company spokesperson Drew Pusateri called the FSU shooting a tragedy but said ChatGPT is not responsible. The company has said it trains its models to refuse requests that could meaningfully enable violence and that it notifies law enforcement when conversations suggest an imminent and credible risk of harm to others, with mental health experts helping assess borderline cases. Those statements preview OpenAI's likely defense posture: that its safety operations meet a reasonable standard of care and that no chatbot output caused the harm. The contested question will be whether the discovery record supports that account.

What courts will likely struggle with

A few things worth watching that do not fit neatly into the public framing of the cases:

Causation will probably be harder than any of the doctrinal questions. Proving ChatGPT contributed to or caused either shooting — as opposed to being one element in a complex set of human and contextual factors — is the kind of evidentiary problem that has sunk product-liability claims in less novel contexts. Even if the doctrinal questions resolve in plaintiffs' favor, causation may be where the cases actually get decided.

Courts may avoid the First Amendment question by ruling narrowly on other grounds. Federal courts generally prefer to dispose of cases on the narrowest available theory. If a court can decide the early motions on Section 230 or product-liability classification grounds, it likely will, leaving the First Amendment doctrine for another day. That preference helps plaintiffs in the short term and creates uncertainty for AI companies looking for clear rules.

The discovery records will look different in the two case sets. Tumbler Ridge relies on specific pre-attack internal flagging allegations sourced to Wall Street Journal reporting. If discovery confirms that record — internal communications, escalation memos, threat-detection logs — those cases become substantially stronger. The Florida case depends less on pre-shooting internal records about a specific user and more on the chat logs themselves combined with documented system-wide safety lapses. Discovery there will focus on the model's design, training data, safety-testing timelines, and decisions to ship versions despite flagged concerns.

Whether these cases consolidate matters more than people think. The Tumbler Ridge cases are in the Northern District of California; the Florida case is in the Northern District of Florida. Different plaintiffs' teams, different jurisdictions, similar underlying defendant. If additional U.S. cases land, the question of consolidation — informal cooperation, MDL coordination, or parallel tracks — becomes more concrete. Watch for any signal of formal coordination; it changes settlement leverage.

What this means for other AI companies

The outcome of the OpenAI lawsuits will shape the legal exposure of every company building or deploying AI systems with public-facing user interaction. If product-liability and process-failure framings survive early dismissal, AI developers, chatbot platforms, and the companies integrating those systems into products will face pressure to demonstrate not just technical safety, but legally defensible safety operations: documented escalation procedures, clear account-enforcement protocols, retained internal safety reviews.

The standard is not going to be whether a company has a safety team. It is going to be what the safety team did with what it knew.

What to watch next

The next major filings will be motions to dismiss, where OpenAI will argue plaintiffs have not established a legally cognizable duty, sufficient causation, or a viable product-liability theory in the face of Section 230 and First Amendment defenses. If any claims survive, discovery in each case will focus on internal safety reviews, escalation policies, account-enforcement records, threat-detection system outputs, training-data curation, safety-testing timelines, and what OpenAI personnel knew before each of the underlying shootings.

Bottom line

The two shooting case sets are best understood as a matched pair pulling in different directions. Tumbler Ridge is the specific-knowledge case, built on an allegation that the company was warned about this user and did nothing; Florida is the systemic-negligence case, built on the chat record plus a documented pattern of safety shortcuts, with an added entrustment theory designed to sidestep the hardest classification question. Each is likely to be narrowed or resolved at the motion stage rather than at trial, the ordinary path for novel tort theories. The early rulings will set the leverage for both, and for cases not yet filed — which is why the motions, not the eventual verdicts, are the thing to watch.

Sources and further reading

• Lawsuit Informer: OpenAI Lawsuits — ChatGPT Wrongful Death, Self-Harm, and Mass Shooting Claims
• Lawsuit Informer: OpenAI Sued Over FSU Mass Shooting — Inside the Chabba Lawsuit
• Lawsuit Informer: OpenAI Sued Over Teen's Overdose Death — Inside the Turner-Scott Case
• Joshi v. OpenAI Foundation et al. (Complaint, N.D. Fla., May 10, 2026)
• Reuters: Family of Florida mass shooting victim sues OpenAI in U.S. court
• WLRN: Family of FSU shooting victim files lawsuit naming gunman and OpenAI
• Reuters: Families of Canadian mass shooting victims sue OpenAI, CEO Altman in U.S. court
• Associated Press: Families of Canada school shooting victims sue OpenAI over shooter's use of ChatGPT
• Wall Street Journal: OpenAI sued by seven families over mass shooting suspect's ChatGPT use
• FindLaw: Garcia v. Character Technologies, Inc.
• Social Media Victims Law Center: Garcia v. Character Technologies case summary
• Florida Attorney General: Criminal investigation announcement involving OpenAI and ChatGPT

Educational commentary only. Not legal advice. No attorney-client relationship is created.