Users have taken to social media over the past two weeks after discovering that the chatbot "ChatGPT" has been using English profanity unusually often.
اضافة اعلان
A recipe generated by ChatGPT and shared by Business Insider included an expletive when describing an amount of "thinly sliced green onions." In another screenshot from a conversation regarding an outage affecting Microsoft's code hosting platform GitHub, the chatbot also employed profane language.
The surge in profanity usage earlier this month coincided roughly with OpenAI's transition from the GPT-5.5 AI model to the GPT-5.6 Luna model as the default in ChatGPT for free users, according to a report by Business Insider.
No clear consensus exists regarding why this shift triggered a wave of profanity, and OpenAI declined to comment.
As OpenAI and its competitors roll out new AI models, companies have sought to make them entertaining to converse with, human-like, and equipped with distinct personalities.
Different models come with various traits that bring them closer to human speech. ChatGPT has long occasionally used profanity in response to user complaints or to emphasize specific points, but users now observe that the chatbot has grown significantly bolder in its use of expletives.
The rate of profanity usage appears to vary between users. Some individuals on the ChatGPT forum on Reddit noted that they assumed the chatbot was mimicking their own speaking style.
Alexis Boger, who uses the chatbot for work and recipe ideas, told Business Insider: "I joked with a friend that I broke ChatGPT because it started mirroring my profanity-laced style." For instance, Boger noticed ChatGPT using the phrase "big ass kolaches" in a recipe suggestion.
Boger added that the profanity felt a bit excessive initially before appearing to taper off.
Meanwhile, a Reddit user told Business Insider that, as an Australian, they felt "these changes made conversations with ChatGPT feel much more natural."
Al Arabiya
