ChatGPT calls user the N-Word out of nowhere during music search

A Reddit user says ChatGPT casually called them the N-Word while helping track down a metal song, and the post quickly caught fire on r/ChatGPT. The thread is titled “My chatgpt said the N-Word,” which tells you pretty quickly why it took off.

In the chat the user shared, they were trying to remember a song from a lyric snippet. Then the bot suddenly used the slur, and when the user pushed back, it did not just apologize and move on. It tried to explain the word away as casual slang, which honestly made the whole thing look even worse.

Redditors in the comments latched on to this. Many joked that they try to bait chatbots into saying awful things all the time, but this one hits differently, at least on its face, because the prompt was about music, and the reply went off the rails anyway.

Interestingly, this isn’t even the first time this seems to have happened. A separate February Reddit post was titled “Chat Gpt Saying the n word with no prompt,” and another thread from March 2025 claimed ChatGPT had used the term “casually” without the user prompting it to.

That said, this does not prove a broad bug or some new break in OpenAI’s filters. It does show this kind of failure has surfaced in public more than once, which is not great for a product that is supposed to be careful around slurs and harassment.

OpenAI is hardly alone here. xAI’s Grok has found itself in hot water a number of times for blurting out slurs and even calling itself MechaHitler in the past.

That is where these “more natural” AI personalities enter a gray area. If the model keeps trying to match the room, sooner or later it is going to match the worst part of it. For instance, last year, WIRED reported that despite OpenAI designing ChatGPT 5 to be safer, it still threw out gay slurs.

Similarly, Vice also reported last year that people were already fixated on trying to get ChatGPT to say the N-Word, so this is not some obscure edge case companies just discovered yesterday.

Still, that is what makes this post stick. The user was not trying to jailbreak anything. They were looking for a song, and ChatGPT, at least according to the exchange they shared, somehow decided that “bro” mode was the move and drove straight into a wall.

Featured image generated with AI

We stand out from the tech-media crowd because we break news stories; we mainly bring you stuff that you won’t find anywhere in the mainstream tech media. Our stories have been picked up by some of the world’s most popular websites and media outlets—more info is available here.