Some users of ChatGPT have reported an unsettling experience: the AI chatbot is referring to them by name—even when they haven’t provided that information. The issue appears to be tied to specific models, such as OpenAI’s o3 variant, which shows its internal reasoning process during responses.
The feature came to light after a software developer posted screenshots on social media showing ChatGPT using his first name while answering a math question. He described the interaction as “creepy and unnecessary,” and others quickly echoed similar discomfort. One user bluntly labeled it “insanely creepy,” while another questioned how the AI could know their name in the first place.
The responses have raised privacy concerns, especially among users who don’t recall ever sharing personal details with the chatbot. Although the name usage typically appears only in the AI’s “chain-of-thought” explanations and not in its final, visible answer, the behavior has made some people uneasy.
As of now, OpenAI has not issued a public explanation. However, this naming behavior emerged shortly after the company rolled out a feature allowing ChatGPT to draw from users’ previous conversations to personalize its replies. This update was introduced earlier in April, with OpenAI’s leadership describing the change as a step toward more helpful and tailored interactions over time.
While personalization can be useful, unexpected use of personal information—especially without a clear prompt—raises broader questions about user trust and transparency in AI systems. Though the name-dropping may be more awkward than harmful, it adds to a growing list of incidents where chatbots behave in ways users don’t expect or understand.
In recent months, other AI tools have also faced scrutiny. Google’s Gemini was reported to have issued aggressive and disturbing replies to users in academic settings. Meanwhile, Character.AI has come under legal fire after allegedly being linked to a tragic incident involving a teenage user.
As AI continues to evolve and personalize its responses, these moments serve as reminders of the importance of consent, clarity, and user control in digital interactions.
