Quick question: Does the Gemini 2.5 integration in Cursor handle privacy like Google AI Studio (where data can be used for training) or like Gemini Advanced (where privacy is respected)?
I know Cursor itself doesn’t collect private data, but I’m specifically concerned about how the Gemini 2.5 integration works when processing my code.
Thanks for your response, but after reviewing the Gemini API Terms of Service, I’m still not clear about how my data is being handled.
The TOS clearly distinguishes between two very different privacy approaches:
Unpaid Services: Google uses submitted content to improve their products and services, human reviewers may access data, and you shouldn’t submit sensitive information.
Paid Services: Google doesn’t use prompts or responses to improve their products, and data is only logged temporarily to detect policy violations.
My question is specifically: Is Cursor using the paid or unpaid version of the Gemini API? This makes a significant difference in how my code would be handled from a privacy perspective.
Can anyone from the Cursor team clarify this important distinction?
Sorry for the delayed reply, but yes, all the model in Cursor including the Gemini models have our full privacy agreements in place, ensuring both us and Google do not store or train their models using the data we send in order to generate AI responses.
Perfect, thank you very much for confirming that privacy is maintained with the Gemini models and that the data is not used for training. I appreciate the clarification.