What us PrivateGPT?
The launch of PrivateGPT, a “privacy layer” for large language models (LLMs) like OpenAI’s ChatGPT, was announced by data privacy provider Private AI. With the help of the new tool, sensitive information and personally identifiable information (PII) from user prompts can be automatically redacted.
Which are the things that PrivateGPT do?
In order to prevent OpenAI from accessing sensitive information, PrivateAI utilises its own AI system to redact more than 50 different categories of PII from user prompts before they are transmitted to ChatGPT. By replacing the PII with placeholder data, users are still able to query the LLM.
How other companies do the same?
The attempts by PrivateAI and Cado Security to tack on extra privacy features to established LLMs demonstrate the importance of data anonymization approaches for businesses wishing to use ChatGPT while limiting their exposure to third parties.