Users have reported a sudden and unexplained increase in the use of profanity by OpenAI's ChatGPT, with many noting the frequent appearance of the 'f-word' in responses.
The phenomenon has sparked speculation about potential changes in the model's safety filters or training data following recent updates to the GPT-5.6 Luna architecture.
OpenAI has not yet issued a formal explanation, but the trend has led to widespread discussion on social media regarding the model's evolving conversational tone and guardrails.