Exercise caution when using Google’s GenAI apps like Gemini since your conversations may be read, labeled, and processed by human annotators to enhance the service.
In a support document, Google discloses the collection, retention, and processing of data from users of its Gemini chatbot apps for web, Android, and iOS.
Google states that conversations with Gemini are stored for up to three years, including related data like language usage, device information, and location. However, users can control this data retention to an extent.
By disabling Gemini Apps Activity in Google’s My Activity dashboard, future conversations will not be saved for review under a Google Account. Additionally, specific prompts and conversations can be manually deleted from the Gemini Apps Activity screen.
Even when Gemini Apps Activity is turned off, conversations will still be stored for up to 72 hours for safety and security reasons, according to Google.
It’s important to avoid entering any sensitive or confidential information into Gemini conversations since human annotators may review and process them.
While Google’s data policies are similar to those of its competitors, the privacy implications of GenAI data retention have led to scrutiny from regulators such as the FTC and the Italian Data Protection Authority.
As organizations become increasingly aware of GenAI privacy risks, a survey from Cisco revealed that 27% of companies have banned the use of GenAI altogether.
While GenAI products geared towards enterprises do not retain data for model training or any other purposes, consumers remain exposed to privacy risks.
