I’m a big fan of this channel and there’s some good content for using new AI tools in our industry. If you aren’t using gpt yet for AT LEAST emails and formatting stuff then you are not as efficient as you could be.
I’m posting this to help foster growth and discussion so feel free to jump in.
Yes, generally if you have access to related information (PII, material non public) then they will restrict rather than take the risk until a full review can take place.
I suppose an all out ban is an easier solution than security training for people who think its ok to put your password on a sticky note on your monitor. If I’m in charge of infosec for a huge organization I’d seriously consider that.
I work at a small firm and do freelancing on the side so no such restrictions. That said, I am careful about putting sensitive data in gpt. So I’ll only feed it numbers and no identifying data like names addresses ssn/ein etc. A sufficiently advanced AI with enough raw data might be able to identify my clients from the anonymized inputs I feed it…but I’m not too worried about that…yet…