Google has confirmed an AI update for all Android users, raising concerns about privacy as its Gemini AI accesses content across various apps. The company’s 2 billion Gmail users also face increased AI integration, which involves using cloud-based AI to process personal and sensitive data.
Recently, Google sent an email indicating that Gemini AI would access apps like Phone, Messages, and WhatsApp even when Gemini Apps Activity is disabled. This caused widespread confusion and alarm, with users fearing unwanted data access.
Google has since clarified that the update aims to improve user convenience. Android devices will use Gemini for tasks such as sending messages and making calls regardless of the Apps Activity setting. However, when Gemini Apps Activity is off, messages and chats are not reviewed or used to train AI models. Previously, keeping this activity off would disable such integrations altogether.
Despite this clarification, interactions with Gemini on Android are temporarily stored for up to 72 hours, even if activity tracking is disabled. Privacy experts warn this short-term data retention still raises risks, especially as Gemini gains deeper access to private communications.
The integration of AI in messaging echoes similar moves by WhatsApp, which recently added AI-generated summaries for chat threads. While these features enhance functionality, they also spark concerns about user awareness and data security, particularly for younger users with limited privacy knowledge.
Google is also rolling out Gemini for educational use, supporting lesson planning and real-time feedback. This expansion prompts debate on AI’s role in education and the potential impact on learning privacy.
According to a recent Gen AI and LLM Data Privacy Ranking report by Incogni, most users are unaware of the complex privacy challenges presented by AI integration. The study highlights risks of unauthorized data sharing and personal information exposure as AI becomes embedded in daily workflows. Google ranks better than Microsoft and Meta in data transparency and model training practices, although privacy concerns remain significant.
TechRadar notes that Gemini for Education follows strict data protection policies, ensuring student data is neither used to train AI models nor reviewed by humans, aligning with Google Workspace for Education standards.