Chrome removes AI privacy wording, Google says data still stays on-device

Google has just removed wording related to how Chrome’s on-device AI models handle data – but the company says nothing has actually changed.

In Chrome version 147, the "On-device AI" message previously said: "To power features like scam detection, Chrome can use AI models that run directly on your device without sending your data to Google servers. When this is off, these features might not work."

But this has been changed in Chrome 148.0.7778.97, which has just started rolling out, with the phrase “without sending your data to Google servers” removed.

The change was noticed by Reddit users and was quickly picked up by privacy researchers and developers.

Computer scientist and lawyer Alexander Hanff said that there are only three reasons why something like this could happen:

1. The claim was inaccurate when it was made, meaning data was being sent to Google servers without disclosure.
2. The model will now start sending user data to Google's servers.
3. Google decided that the original text was too rigid and hard to defend in court.

Hanff adds that none of these are “good” explanations, ranging from direct breaches of the GDPR, the ePrivacy Directive, and consumer protection law.

He asked Google direct questions: “Why was the sentence ‘without sending your data to Google servers' removed from the on-device AI description in Chrome's Settings UI? Was the previous text inaccurate? Has the architecture changed? Was the wording withdrawn on legal advice because Google was unwilling to defend it as a representation?”

The Register reports that a Google spokesperson said that nothing has, in fact, changed: "This doesn’t reflect a change to how we handle on-device AI for Chrome. The data that is passed to the model is processed solely on device."

However, Google said that in some instances, websites that use Gemini Nano in Chrome will be able to see the model’s inputs and outputs, such as during interactions via the Prompt API. This will be handled by each website’s individual privacy policy.

The conversation on this topic is live. Join in the discussion.

According to Google, the wording was changed to avoid potential confusion.

Just this week, it was revealed that Google Chrome is silently installing a 4GB AI model called Gemini Nano on user devices without consent. Hanff said that the move can violate privacy laws and negatively impact the environment through CO2 emissions, as mass distribution can generate between 6,000 and 60,000 tonnes of CO2-equivalent emissions.

The timing of the rollout, along with the privacy wording change, sparked speculation that Google could eventually begin collecting on-device prompts and responses.

Users who are currently on version 147 will see the old text under Settings > System > On-device AI, while those who updated to the most recent version 148 will already see the new text.

Unlock more exclusive Cybernews content on YouTube.