This analysis describes what Google Gemini's agreement states, permits, or reserves. It does not constitute a legal determination about enforceability. Regulatory applicability and practical outcomes may vary by jurisdiction, enforcement context, and individual circumstances. Read our methodology
It establishes that user data is not confined to operational service delivery but also feeds into the training and development of the underlying AI and machine-learning models.
Interpretive note: The excerpt references 'these uses' without specifying them within the quoted text; the full scope of those upstream uses is not established in this clause alone, which limits confidence in the complete picture.
The updated privacy notice discloses a new calling feature that allows Gemini to make calls on behalf of users on eligible Pixel devices. Before each call, Gemini will display the information it plans to share and require user confirmation; sensitive data like credit card numbers and passwords will not be shared. Call transcripts are stored on Google servers and used to generate summaries and display calling history, but audio recordings are not stored on Google servers or used to train AI models. You can delete your calling activity, transcripts, and recordings at any time through your Gemini Apps Activity or directly in the Phone by Google app.
View change record →The updated privacy notice discloses that when using the new Speak-to-Window feature on macOS, Gemini captures audio from your microphone along with your active window content to process your instructions. Transcripts of your voice instructions are retained according to your Keep Activity setting and used to improve Gemini, including through human review. The audio data itself is not stored. You can control transcript retention by adjusting your Keep Activity setting in Gemini.
View change record →The updated notice establishes that custom third-party connected apps integrated with Gemini are not monitored or secured by Google, and warns that connecting them may expose your data, passwords, devices, and accounts to unauthorized access. The notice also discloses that Gemini Spark on macOS can interact directly with local files you add to connected folders, including editing and organizing them. Additionally, the notice introduces temporary chats as an option that are not used to improve Google AI with human reviewers, though Google states it still uses chats to respond to you and help protect Google, users, and the public even when Keep Activity is off or temporary chats are used. You can start a temporary chat to opt into this alternative chat mode, and you can manage Gemini Spark settings to control access to local files.
View change record →Data you provide to Google Gemini may be used to train or improve the generative AI models and machine-learning technologies behind the service.
How other platforms handle this
Comply with applicable laws, lawful requests, and legal process, such as to respond to subpoenas or requests from government authorities
Any personal information we collect from you to verify your identity in connection with your request will be used solely for the purposes of verification.
We may use information about you, your activity, content, media uploads and related data in your account to prevent, detect, investigate and address safety, security, fraud and abuse risks, and to develop our algorithms and models to identify violations...
"These uses extend to the generative AI models and other machine-learning technologies powering our services.Excerpt from Google Gemini's Gemini Apps Privacy Notice
We read the privacy policies and terms of service of 38 AI platforms. Here is what they say about training, retention, arbitration, and liability.
Get the weekly coverage letter
Companies change their terms without announcing it. We capture governance documents, compare each new version with the last, and publish what changed, with the source text behind it. One email a week on what we found and what we could not reach.
It establishes that user data is not confined to operational service delivery but also feeds into the training and development of the underlying AI and machine-learning models.
Data you provide to Google Gemini may be used to train or improve the generative AI models and machine-learning technologies behind the service.
ConductAtlas has identified this type of provision across 272 platforms. See the full comparison.
No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by Google Gemini.