If you’ve noticed an uptick in profanity from ChatGPT, you’re not alone. Over the past week, users have taken to social media to express their surprise at the chatbot's newfound penchant for vulgar language—though most reactions have been positive.
One notable instance occurred when ChatGPT offered a recipe to Business Insider that included the phrase "thinly sliced green onions" accompanied by an unexpected expletive. In another interaction revolving around a technical issue with Microsoft's GitHub, the chatbot humorously remarked, "GitHub is having a shitfit."
This shift in language occurs around the same time OpenAI transitioned from the GPT-5.5 model to GPT-5.6 Luna for free users. The connection between this update and the increase in foul language remains unclear, as OpenAI has opted not to comment on the matter.
As companies like OpenAI enhance their AI models, they strive to create chatbots that are engaging, relatable, and fun. Each model has its unique character traits, and historically, ChatGPT has occasionally sprinkled in mild swearing during conversations. However, it appears that the chatbot has recently cranked up the intensity.
The frequency of these expletives seems to vary across different interactions. Some users in the ChatGPT subreddit speculate that the chatbot adapts to their speech patterns. Alexis Boger, a user who utilizes the chatbot for both work and recipe generation, humorously noted that she felt she had "broken" it when it responded to her with phrases reflecting her own informal language, such as "big ass kolaches." While initially finding the swearing a bit too frequent, she mentioned it subsequently leveled off.
Many users have embraced the more colorful language that ChatGPT is using. One Redditor expressed their excitement, saying they were "ecstatic" about the shift. An Australian user also shared that the increased swearing made the interactions feel much more natural for him. He pointed out that while he noticed a flurry of profanity initially on August 9, it has since decreased.
Timothy Hartman, who employs OpenAI’s tools for coding, adjusted his ChatGPT’s settings for a more "casual and opinionated" tone, encouraging it to be less reserved with its language. He observed the rise in cursing this month and shared examples that showcased the chatbot's boldness, even using strong language during technical discussions.
Hartman reflected on how the chatbot's use of profanity evoked a saying from his father: "Profanity is the ignorant person's attempt to express themselves forcefully," alluding to the nuances of communication.
If you have any insights or stories to share, feel free to reach out to this reporter via email or text using secure channels.
