With over three billion individuals depending on Google Chrome, any undisclosed modifications to the software are particularly concerning. Recently, security analyst Alexander “The Privacy Guy” Hanff disclosed that Chrome has been covertly downloading a four-gigabyte AI model onto users’ devices without their consent. This alarming revelation has sparked significant outrage across the internet.
Hanff identified a file titled “weights.bin” located in a directory named “OptGuideOnDeviceModel.” This file includes the parameters for an AI model known as Gemini Nano, which is intended to operate locally on devices rather than depending on cloud solutions. “Chrome did not ask,” Hanff remarked in his blog, highlighting that users remain entirely unaware of this installation. If users attempt to remove the file, Chrome automatically reinstalls it.
The consequences of this hidden installation extend beyond mere storage concerns, raising issues regarding device performance and user privacy. As skepticism towards AI technologies increases, Google’s lack of transparency exacerbates these worries. The corporation has yet to issue any public comments regarding the situation and has disregarded media requests for clarification.
Hanff also raised a troubling environmental concern, estimating that the rollout of this AI model could produce “between six thousand and sixty thousand tonnes of CO2-equivalent emissions,” contingent on the number of devices involved. This statistic adds to user annoyance, as many have long disliked unsolicited software installations.
Reactions online have been rapid and fervent. A Reddit user expressed discontent: “AI/Climate aside, my real issue is that Google installs anything without my consent no matter the size. Thankfully I don’t use Chrome; it’s Firefox for me.” Others speculated that Google might be attempting to artificially inflate user engagement statistics for its AI offerings. “All of this only so they can show the market that people really are ‘using’ their AI crap,” another commenter noted.
Another user encapsulated the prevailing sentiment: “AI is so s*** that companies have to force it down people’s throats. They know nobody asked for this.” The backlash persisted, with one user claiming, “There were few things Google could do to force users off of Chrome faster than this.”
In his blog, Hanff clarified that the AI model downloads when Chrome’s default AI features are turned on. “On any machine that meets the hardware requirements, Chrome treats the user’s hardware as a delivery target and writes the model,” he detailed. To prevent Chrome from reinstalling the model after it has been deleted, users must manually switch off AI features in the browser settings. “That is the true definition of malware,” a user on X commented, mirroring the growing discontent.
In addition to environmental concerns, Hanff suggested that Google’s behavior might breach European Union data privacy regulations, including the General Data Protection Regulation (GDPR), which mandates strict consent for the use of personal data. This could lead to significant legal repercussions for the tech giant.
Chrome is not the only browser under scrutiny for its AI features. Mozilla has also faced criticism and has pledged to introduce a “kill switch” to disable all newly added AI functionalities in Firefox. This response highlights the industry’s increasing recognition of user apprehensions regarding AI.
Fortunately, alternatives to Chrome exist that emphasize user control and privacy. Vivaldi, for instance, has embraced a distinct approach. In an August blog entry, CEO Jon von Tetzchner reaffirmed the company’s dedication to user autonomy, stating, “We will continue building a browser for curious minds, power users, researchers, and anyone who values autonomy.” He also noted that Vivaldi would only incorporate AI if it aligns with their values, safeguarding user rights and intellectual property.
As this situation unfolds, the ramifications of Google’s actions are prompting users to reevaluate their browser options and the importance of consent in an increasingly AI-oriented landscape. Will this incident lead to a change in browser preferences, or will convenience continue to outweigh privacy issues? Only time will tell.
Via: futurism.com
