OpenAI Will Now Tell Parents If Their Teen Got Banned From ChatGPT for Violent Threats

OpenAI is rolling out a new alert that tells parents when their teenager's ChatGPT account gets shut down for violating the company's rules against violent threats or acts of violence.

Parents who've linked their accounts to a teen's will get a notification about an "important update." Clicking through gives them the full picture: the account was deactivated for violating OpenAI's policy on Acts of Violence.

This builds on parental controls OpenAI launched last year, which already let parents set usage hours, dial back sensitive content, and get alerted if the system flags signs a teen might be considering self-harm. The violence notification is new. Before this, a parent could have no idea their kid got banned for threatening someone, unless the kid told them.

Why Now

OpenAI built this feature with Moonshot, a firm that specializes in tracking and countering online violence. Moonshot founder Vidhya Ramalingam said notifying a parent when a serious concern comes up, with a path to more context, is "a critical first step in giving families the chance to step in early and seek help." She also called it one of many measures still needed to keep young people safer on the platform.

The timing isn't a coincidence. OpenAI has been under pressure since it came out that the suspect in the 2025 mass shooting in Tumbler Ridge, British Columbia, had used ChatGPT. OpenAI had already banned the suspect's account for policy violations before the shooting, according to Engadget, but the company never flagged it to Canadian authorities. Canadian officials summoned OpenAI over the failure, and CEO Sam Altman apologized publicly for the oversight.

A private company had information suggesting a user was dangerous and sat on it. Nobody outside OpenAI knew until after people were dead. Whatever you think of AI chatbots generally, that's a legitimate failure, and it's reasonable that regulators and the public expect something to change.

What This Doesn't Fix

Parental notification only works if a parent has linked their account to their teen's in the first place. OpenAI hasn't said what percentage of teen users actually have linked parental accounts, and Engadget's coverage doesn't address it either. A notification system is worthless for the kids whose parents never set it up, which is presumably a lot of them.

This feature tells a parent after the fact that an account got banned. It's not intervention, it's an after-action report. Whether that's enough to prevent another Tumbler Ridge is an open question nobody can answer yet, including OpenAI.

There's also the question of law enforcement notification, separate from parental notification. The Tumbler Ridge case was a failure to alert authorities, not parents. This rollout addresses the parent side. Whether OpenAI has changed its policy on when it proactively contacts police or Canadian and U.S. authorities about accounts flagged for violent threats is not addressed in this announcement.

The Rest of the Update

OpenAI bundled a couple of other changes into the same rollout. Parents can now flip on "Study Mode" directly from the parental controls panel, which makes ChatGPT give hints before answering homework questions instead of just handing over the answer. And teens who spend a long stretch in a single ChatGPT session will start seeing more frequent break reminders.

Both are minor next to the violence notification, but they fit a pattern. OpenAI is trying to look like a responsible actor for teen users after a body count forced the issue. Critics on child-safety issues will say this is reactive, not proactive, arriving only after a shooting and a government summons rather than being built in from the start.

OpenAI hasn't said whether this notification system will be reviewed by outside auditors or whether Canadian authorities consider the company's response adequate going forward. That's the next thing worth watching: whether Ottawa treats this as a resolved matter or keeps pressing for stronger disclosure requirements when an AI company's own product flags a user as a violence risk.