Google’s AI assistant Gemini has sparked fresh Gmail AI privacy concerns after the company expanded data collection settings and integrated more AI-powered features into its services. Privacy advocates argue that many users may not realize how much information Google’s AI can access or process by default.
Although Google says it does not train its core AI models on personal Gmail content, critics believe the company’s latest approach gives users too little control over their data.
Gemini Can Process Gmail Content
Gemini can access Gmail messages to summarize email threads, answer questions, and help users manage their inboxes.
To deliver those features, the AI processes email content during requests. Google says this information remains private and that Gemini does not retain email data or use personal Gmail messages to train its foundational AI models.
Even so, privacy advocates point out that the AI still needs access to inbox content to perform those tasks.
Google Enables New Data Collection by Default
Google has also started rolling out a new setting called Search Services History.
The feature is enabled by default for many users and collects activity across Google Search services. That data may include images, audio and video recordings, uploaded files, and user interactions. Google uses this information to improve its generative AI systems.
Critics argue that many people will never discover the setting because they rarely review their privacy preferences.
Gemini Highlights “Opt-Out Fatigue”
Privacy-focused email provider Proton Mail has criticized Google’s approach and encourages Gmail users to move to its encrypted email platform.
When asked about the issue, Gemini acknowledged that Proton’s warning also serves as marketing for its competing service. However, the chatbot also discussed what privacy advocates describe as opt-out fatigue.
According to Gemini, companies often enable new AI features by default because many users never change the settings. The chatbot suggested that reviewing privacy controls gives users greater control over how companies process their data.
Google Rejects Claims of AI Training on Gmail
Google continues to distinguish between AI assistance and AI training.
The company says Gemini can process Gmail messages to provide features such as summaries and smart replies. However, it maintains that personal Gmail content does not train Gemini’s foundational AI models and that the service does not retain email data after completing requests.
For many users, the debate comes down to trust. Some people accept AI processing in exchange for added convenience, while others prefer to disable the features entirely.
How to Limit Gemini’s Access
Users who want to reduce Gmail AI privacy risks can disable Gmail’s Smart Features and Personalization setting.
Google also allows users to review the new Search Services History option through their Google Account. There, users can disable media collection and adjust other AI-related privacy settings.
As Google expands Gemini across its products, privacy experts recommend reviewing account settings regularly instead of relying on default options.
0 responses to “Gmail AI Privacy Concerns Grow as Gemini Expands Data Access”