A number of concerns surrounding privacy have surfaced since the arrival of ChatGPT-3.5 last November. The company that created ChatGPT - Open AI - has been clear from the start about its intention of improving the tool by making it free to users. (ChatGP- 3.5 is free, ChatGPT-4 costs $20/month.) So, ChatGPT-3.5 is in something of a beta phase with current - and many unwitting - users identifying the bugs for free. Information provided by the user is then available for training purposes. After meeting criticism, however, Open AI has since offered a way for users to opt out of having their conversations used for training purposes. All conversations are held for 30 days. Additionally, Open AI will make some information available to unspecified third parties. 

Ostensibly the tool is not available for children under the age of 13. However, there are no age verification checks or parental permissions required when registering. Because of the law COPPA - Children's Online Privacy Protection Act - websites are not authorized to collect any personal information from children under the ages of 13. Initially, Open AI recommended that children between the ages of 13 and 18 use the tool under parental supervision/permission. That has since been relaxed.

Because of the risk of inadvertently supplying the bot with sensitive, business or personal information, companies are already taking the step of banning its use. Three of these businesses include Samsung, Goldman Sachs and Apple. Professors opting to use ChatGPT for assignment purposes should make students aware of the privacy issues and caution students to be careful about what kinds of  information they provide the tool when using it. They may also want to offer students the opportunity to opt out if they choose. Because the tool is so new, universities have not had time yet to develop their own policies about its use.

The links included in "Associated Links for Privacy Concerns" includes Open AI's privacy policy for ChatGPT and a separate article on its rationale. Note that Open AI's privacy policy is in a fluid state as it reacts to bans, concerns and proposed regulations.