OpenAI Flagged Canada Shooter’s Account Months Before Attack

OpenAI employees flagged disturbing messages from the Tumbler Ridge shooter months before he carried out his deadly attack, but their warnings didn’t reach the right people in time.

That’s a big part of why openai canada shooter keeps coming up in the conversation. The AI company had banned the shooter’s ChatGPT account back in June 2025. That’s a full eight months before the tragedy that would rock the small B.C. Community.

This is honestly pretty damning. OpenAI’s safety team knew something was up. They saw the red flags in this guy’s chat logs and took action to boot him from the platform. But somehow that intel never made it to Canadian authorities until after the shooting happened.

The Timeline That Raises Questions

Here’s what we know about the sequence of events. In June 2025, OpenAI’s content moderation systems or human reviewers spotted concerning messages from the future shooter’s account. The company made the call to ban him from accessing ChatGPT.

Fast forward to the shooting itself. Only after the attack did OpenAI reach out to the RCMP to share what they had on file about the perpetrator’s online activity. For more on openai teams indian, check out OpenAI Teams Up with Indian Universities for Massive AI Push.

B.C. Premier David Eby called the allegations “disturbing” when asked about the reports. That’s putting it mildly.

“If true, this raises serious questions about information sharing between tech companies and law enforcement.”

There wasn’t any mention of concerns from OpenAI in the immediate aftermath of the shooting. Radio silence from the company for at least a day after the tragedy.

What Was in Those Messages?

We don’t have the full details of what exactly the shooter was asking ChatGPT that got him flagged. OpenAI hasn’t released specifics about the concerning content, and honestly, they probably shouldn’t for obvious reasons. For more on canadians pretty done, check out Canadians are pretty done with America right now, poll shows.

But we can make some educated guesses. The company’s safety systems are designed to catch users trying to get help with violence, weapons, or harmful activities. If they banned his account, he was likely pushing boundaries that their AI wasn’t supposed to cross.

Thing is, ChatGPT has gotten pretty good at refusing to help with sketchy requests. You ask it how to make explosives or plan an attack, and it’ll shut you down fast. But determined users sometimes find workarounds or phrase things in ways that slip past the filters.

The Information Sharing Problem

This whole situation shows a massive gap in how tech companies handle potential threats. OpenAI spotted something concerning enough to ban the user, but there’s apparently no standard protocol for alerting authorities about credible threats.

Should there be?

That’s a complex question. Privacy advocates would rightfully freak out if AI companies started routinely sharing user data with police. But when someone’s using your platform to potentially plan violence, the calculus changes.

Other tech giants like Google, Meta, and Microsoft all have their own policies for handling concerning content. Some proactively report to the National Center for Missing & Exploited Children when they spot illegal material. But there’s no consistent approach for potential violence threats.

The Technical Challenge

From a technical standpoint, this isn’t an easy problem to solve. AI systems generate millions of conversations daily. Most concerning messages are just people venting, asking hypotheticals, or writing fiction.

Separating genuine threats from edgelord nonsense requires human judgment. And even then, you’re talking about prediction, not certainty. How many false alarms would authorities get if every tech company started forwarding “suspicious” AI conversations?

But in this case, OpenAI clearly thought the threat was real enough to ban the account.

That should’ve triggered something more than just a platform removal.

What Happens Next

The RCMP investigation into the Tumbler Ridge shooting is still ongoing. They’ll likely be digging deep into the perpetrator’s digital footprint, including whatever OpenAI eventually shared with them.

There’s also going to be pressure on the federal government to establish clearer guidelines for tech companies. When should they report concerning activity? To whom? What’s the threshold?

OpenAI, for their part, will probably face some tough questions about their incident response procedures. They caught the red flag but fumbled the handoff to authorities.

The company has been pushing hard on AI safety lately, hiring former government officials and building out their safety teams. But this case shows there are still major blind spots in how they handle real-world threats identified through their systems.

Look, I get that OpenAI doesn’t want to become the thought police. Nobody wants AI companies reading every message and reporting users to authorities. But when your safety team spots something serious enough to ban an account, maybe don’t just delete and forget.