Google Chrome has come under scrutiny after security researchers discovered that the browser is silently downloading AI models to users’ devices without providing a clear consent flow. The finding has sparked renewed debate about tech companies’ approach to user transparency in the age of AI-powered features.

According to researchers, Chrome has been quietly downloading machine learning models in the background to power various “on-device AI” features — including smart text suggestions, image analysis, and autofill enhancements. The downloads occur automatically during browser updates, with no explicit notification or opt-in prompt presented to users.

静默下载AI模型引发隐私担忧

安全研究人员发现，Chrome浏览器在更新过程中会自动下载多个AI模型文件到用户本地设备，用于驱动所谓的"端侧AI"功能，包括智能文本建议、图像分析和自动填充增强等。整个过程没有明确的用户提示或同意机制。

关键争议点：

  • 缺乏透明通知：用户不知晓AI模型正在被下载到本地
  • 存储空间占用：单个AI模型体积可达数百MB，多个模型累计占用显著
  • 网络资源消耗：后台下载可能影响用户网络体验
  • 数据用途不明确：模型收集和处理的用户数据范围不够清晰

Industry Response

Privacy advocates argue that this practice violates the principle of informed consent. “Downloading AI models that will process user data should require explicit opt-in, not be buried in a browser update,” said one security researcher.

Google has not yet issued a formal response. The company has previously stated that on-device AI processing is more privacy-friendly than cloud-based alternatives, as data never leaves the user’s device. However, critics note that this does not excuse the lack of transparency about what is being installed and why.

This incident adds to a growing list of concerns about how major tech companies deploy AI capabilities without adequate user awareness, echoing similar controversies around Windows Copilot and Apple’s AI features.