Three Tips to Secure Your Privacy When Using AI

In our previous blog we discussed the current state of AI users’ privacy as well as the ethical implications and need for security. This blog will review the steps that you – as a user of the technology – can take to secure your own privacy.  

1. When using AI, avoid sharing personally identifiable information (PII) such as your full name, address, phone number, date of birth, or government identifications. The AI can store these in its memory.  

2. Use temporary chats when possible. You can turn a regular ChatGPT chat into a temporary chat by clicking the dashed speech bubble in the top right corner. These chats retain information only for the duration of that chat window. This prevents data aggregation in which data from past chats are compiled to build a user profile. Do note, however, that temporary chats are not entirely private, though they are less likely to be retained.  

3. Ask “What do you know about me?” This lets you see what information about you the AI has stored. Often it will contain your interests, conversation topics, but also more sensitive information such as your name and occupation (if you have provided it). You can also request, “Forget everything we have talked about in the past,” and the AI will delete these details from its memory.  

These tips will help you maintain control over your data while using artificial intelligence. Next time you use AI, I encourage you to try tip #3. It can reveal a lot!