Chatbots remember you
If you use a chatbot while logged in, you have probably noticed it remembers things about you. It uses those details going forward and keeps learning, and now and then it says something surprising that it picked up from an earlier chat.
OpenAI is open about how this works. Its Memory in ChatGPT help page (2026) says that with memory on, ChatGPT can use details from past chats, saved memories and other sources to personalize answers. It also says memory does not keep every detail, and that ChatGPT decides what is relevant.
Ask it what it remembers
Try this. Open the chatbot you use most and ask, "What do you remember about me?" It will tell you what it thinks it knows. OpenAI's help page suggests the same thing.
When I did that, I was surprised. I do a lot of research for clients, and I often ask questions as if I were the client. A couple of different chatbots had concluded that I worked for those companies, that I was in their industries instead of mine, and in one case that I was the founder.
Why that matters for consultants
That wrong picture colors every answer after it. If the tool thinks you run a manufacturing company, its advice on your own marketing or your own pricing will be shaped by that.
So I went back and corrected it. I told it I don't work for that company, I'm not the founder, and I'm a consultant who was asking questions on their behalf. OpenAI's page says you can correct remembered information, delete it, or use a temporary chat that doesn't create memories.
A small habit worth adding
If you do client research in a chatbot, check its memory every so often. Correct anything that is wrong, and consider a temporary chat or a separate project for each client's work.
AI is a useful research tool for me, but it still needs a person checking it. If you want to talk about how fractionals are using these tools, book a networking chat.
