Technik

OpenAI-Mitarbeiter haben vor Monaten wegen Massenerschießungsverdächtiger Alarm geschlagen: Bericht

22.02.2026

https://www.the-independent.com/news/world/americas/canada-school-shooting-chatgpt-openai-b2924939.html

GetsBetterAfterAFew on 22.02.2026 2:57 a.m.

83% of the dead people in Gaza were citizens, and no one cared, why worry about this marginal amount of people in this shooting? We know Isreal is commiting genocide and noone lifted a finger. I’m not being callous, all life is important, but if people dont give a shit about kids dying in Gaza, why would anyone raise the alarm on a potential shooter? Make the change you want to see …
Mr_Greystone on 22.02.2026 2:59 a.m.

Internal alarms only over several months of time. That’s a pretty quick record for a 3 year old product in advancing societal stress. Pretty sure it’s too late to hit the brakes. If there were, OpenAI likely would have done it months ago when they had these conversation logs. Instead, here we are finding out all about their foul up which messes with us all even if you never used AI in that time. Others share the blame as well, but even OpenAI accepts being the first.
ew435890 on 22.02.2026 3:10 a.m.

My question about this would be how many other people have they had the same alarms go off for, and the person never committed any act of violence? Id be willing to be its a very high number.