Google Accused of Installing Gemini Nano Through Chrome Without Clear Consent: Google is facing fresh scrutiny after reports claimed that its Chrome browser has been quietly downloading an on-device AI model onto some users’ computers without clearly notifying them first.
The controversy centers around Gemini Nano, a lightweight artificial intelligence model designed to run locally on devices instead of relying entirely on cloud servers. While Google promotes the technology as a way to deliver faster and more private AI-powered features, critics argue that many users were never explicitly informed the software had been installed in the first place.
The issue was highlighted by Swedish computer scientist and privacy advocate Alexander Hanff, who claimed Chrome had downloaded the AI model onto supported devices automatically, without requesting direct user approval.
According to Hanff, most users would not even realize Gemini Nano exists on their systems unless they manually searched for its files.
What Is Gemini Nano?
Gemini Nano is Google’s compact AI model built to perform certain tasks directly on phones and computers. Unlike cloud-based AI systems that process requests remotely through internet-connected servers, Gemini Nano handles specific workloads locally on the device itself.
Google has increasingly positioned on-device AI as a major part of its broader artificial intelligence strategy. Running AI locally can reduce response times, improve offline functionality, and lower dependence on cloud infrastructure.
The model powers several AI-driven capabilities, including:
- Writing assistance
- Scam call detection
- Audio summarization
- Smart text suggestions
- Image and screenshot analysis on supported devices
The AI model is separate from Chrome’s AI Mode search features, which still rely on Google’s servers for processing.
Users Reportedly Weren’t Clearly Notified
The criticism is less about the technology itself and more about how it was deployed.
Hanff said Chrome users were not clearly alerted when Gemini Nano was downloaded. Instead, the installation reportedly occurs in the background if a system meets the necessary hardware requirements.
According to reports, the AI model may occupy several gigabytes of storage space on a device.
Users who want to check whether Gemini Nano is present can reportedly search their systems for a folder named “OptGuideOnDeviceModel.” Inside that directory is a file called “weights.bin,” which contains the AI model data.
The folder can typically be found using:
- File Explorer on Windows
- Finder on macOS
- Files on ChromeOS
Hanff argued that most users would never think to search for such files manually, meaning many could remain unaware the model had been installed at all.
Google Says Users Can Disable It
In response to concerns, Google confirmed that Gemini Nano may download automatically on supported systems, but said users can disable the feature through Chrome settings.
A Google spokesperson stated that the model is designed to uninstall itself automatically if a device lacks sufficient processing power, memory, storage capacity, or network resources.
The company also said it introduced controls earlier this year allowing users to turn the feature off entirely.
Once disabled, Gemini Nano should stop downloading or updating on the device.
Google has additionally published support documentation explaining how on-device AI models function within Chrome.
Why Google Is Moving AI Onto Devices
The rollout reflects a broader industry shift toward local AI processing.
Traditionally, advanced AI systems relied heavily on cloud infrastructure, requiring companies to process enormous amounts of data in remote data centers. But as AI usage grows, those server costs are increasing rapidly.
By running some AI workloads directly on users’ hardware, companies can reduce cloud expenses while also delivering faster performance.
Hanff suggested this may be one reason behind Google’s aggressive push toward on-device AI.
Running AI locally allows companies to offer new features without handling every request through expensive centralized computing systems.
At the same time, local AI processing introduces new concerns around storage usage, battery life, system resources, and transparency.
Privacy and Regulatory Concerns
The situation has also sparked discussion around digital consent and privacy regulation, especially in Europe.
Hanff argued that quietly deploying AI software onto personal devices could potentially conflict with transparency requirements under the European Union’s General Data Protection Regulation (GDPR).
He also questioned whether the environmental impact of large-scale AI deployments should be more openly disclosed under Europe’s sustainability reporting rules.
Google has not publicly commented on those specific legal arguments.
Still, the controversy reflects a growing challenge facing the tech industry as AI becomes more deeply integrated into everyday software products. Features once considered optional are increasingly being embedded directly into browsers, operating systems, and mobile devices — often operating quietly in the background.
For users, the debate is becoming less about whether AI exists in their software and more about how much visibility and control they have over it.
Control Your Entire Home in One Sentence? Gemini 3.1 Says Yes | Maya
