What is Gemini Nano and why was it installed on users' devices?

Gemini Nano is an on-device AI model developed by Google, which powers features such as scam detection and text assistance. It was installed without user consent on users' devices to provide these features.

Why is the installation of Gemini Nano concerning for the adult industry?

The blurred lines between legitimate features and malicious intrusions raise concerns for the adult industry, where trust and transparency are paramount.

Google's Chrome Installs AI Model Without User Consent - Privacy Concerns

Q: Where can the Gemini Nano file be found on a device?

The Gemini Nano file, 'weights.bin', is located in a folder called 'OptGuideOnDeviceModel' within Google Chrome's user data folder.

Q: When did the installation of Gemini Nano begin?

The installation of Gemini Nano began at least as far back as April 2026.

Q: How can users remove or turn off Gemini Nano?

Users can turn off and remove the model directly in Chrome settings, a feature that began rolling out in February 2026.

Swedish computer scientist discovers Google's Gemini Nano AI model installation on Chrome without user consent, raising privacy concerns.

Google's Chrome browser has been silently installing a 4GB AI model on users' devices without their consent, sparking concerns about privacy and trust in everyday software. The model, called Gemini Nano, is an on-device AI model that powers features such as scam detection and text assistance, but its installation raises questions about user consent and the blurred lines between legitimate features and malicious intrusions.

What Happened

Alexander Hanff, a Swedish computer scientist and lawyer known as "That Privacy Guy," discovered the issue after noticing that Chrome was installing a 4GB file called "weights.bin" on his device without prompting him. The file is located in a folder called "OptGuideOnDeviceModel" within Google Chrome's user data folder. Hanff found that the model reinstalls itself if deleted, and that users are not notified or asked for consent before the installation.

According to CNET, the installation of Gemini Nano began at least as far back as April 2026, and it is still unclear how many people have been affected. The model requires specific hardware requirements to be installed, but Hanff notes that even if users meet these requirements, they are not informed about the installation.

Background and Context

Gemini Nano is an on-device AI model developed by Google, which allows for features such as scam detection and text assistance without sending user data to the cloud. However, its installation raises concerns about user consent and trust in everyday software. Hanff notes that this is not an isolated incident, citing a previous case where Anthropic silently registered a Native Messaging bridge in seven Chromium-based browsers on every machine where Claude Desktop was installed.

Google has stated that Gemini Nano will automatically uninstall if the device doesn't have enough resources, such as processing power, memory, storage space or network bandwidth. Additionally, users can turn off and remove the model directly in Chrome settings, a feature that began rolling out in February 2026. However, Hanff argues that this is not sufficient, citing concerns about user consent and the potential for malicious intrusions.

Why it Matters to the Industry

The installation of Gemini Nano raises significant concerns for the adult industry, where trust and transparency are paramount. The blurred lines between legitimate features and malicious intrusions can have serious consequences, including compromised user data and reputational damage. Furthermore, the lack of user consent and notification can erode trust in everyday software, making it more difficult to spot malicious activity.

The industry relies heavily on robust cybersecurity measures to protect against threats such as malware and phishing attacks. The silent installation of Gemini Nano highlights the need for greater transparency and accountability in software development, particularly when it comes to user consent and data protection.

What Comes Next

Users can remove the 4GB AI model from their device by uninstalling Chrome or turning off the feature directly in Chrome settings. However, Hanff argues that this is not a sufficient solution, citing concerns about user consent and the potential for malicious intrusions.

The incident highlights the need for greater transparency and accountability in software development, particularly when it comes to user consent and data protection. The industry must prioritize robust cybersecurity measures and ensure that users are informed and notified about any installations or features that may compromise their trust.

Key Facts

Google's Chrome browser has been silently installing a 4GB AI model called Gemini Nano on users' devices without consent.
The model is an on-device AI model developed by Google, which powers features such as scam detection and text assistance.
Gemini Nano requires specific hardware requirements to be installed, but users are not informed about the installation.
Users can remove the model by uninstalling Chrome or turning off the feature directly in Chrome settings.
The incident raises concerns about user consent and trust in everyday software, particularly in the adult industry where transparency and accountability are paramount.

The installation of Gemini Nano highlights the need for greater transparency and accountability in software development, particularly when it comes to user consent and data protection. The industry must prioritize robust cybersecurity measures and ensure that users are informed and notified about any installations or features that may compromise their trust.