Google is facing criticism after privacy researchers revealed that its Chrome browser has been quietly downloading a 4GB artificial intelligence model, known as Gemini Nano, onto eligible computers.
Gemini Nano is a compact version of Google’s AI technology designed to run directly on users’ devices rather than relying solely on cloud-based processing. The model powers a growing range of AI features within Chrome, including content summarisation, text assistance, scam detection, and other on-device capabilities.
According to researchers, the AI model is being installed automatically on compatible systems without a prominent notification or explicit consent prompt. The software is stored as a file named weights.bin and can consume approximately 4GB of storage space.
Google says the model is only downloaded to devices that meet certain hardware requirements and that it may be automatically removed if the system lacks sufficient storage, memory, processing power, or network resources.
The company argues that running AI tasks locally can improve performance, reduce dependence on internet connectivity, and enhance privacy by keeping certain processing activities on the device rather than sending data to external servers.
However, privacy advocates have raised concerns about transparency, arguing that many users may be unaware that a large AI model has been installed on their computers.
The development highlights the growing trend among technology companies to embed AI models directly into consumer devices as they seek to deliver faster and more responsive AI-powered experiences.
