How to Exclude Yourself from Google Search’s Enhanced AI Data Training System

AI models require more than just text data to evolve; they also thrive on varied inputs such as audio and video. If Google can amass a greater quantity and variety of data from its extensive user base, it might be able to innovate more swiftly than its rivals.

Google benefits from a vast user demographic spread across numerous services, enhancing its data collection capabilities. “Google holds a unique position compared to many other companies,” explains Thorin Klosowski, a senior security and privacy advocate at the Electronic Frontier Foundation. “They provide a wide range of services that users have been utilizing for years, leading to a level of comfort and complacency regarding the data that is gathered.” Daily-used apps carry an inherent momentum, meaning that unwanted changes may not be compelling enough to drive users toward alternatives.

The default setting for AI training seems to be opting out across various platforms. This model isn’t necessary. “I believe ‘opt in’ is merely the minimum expectation from these companies,” Klosowski asserts. “They should require users to explicitly choose to activate these features, as a basic courtesy.” To encourage engagement, Google would need to convincingly explain why these features are beneficial if not enabled automatically, Klosowski adds.

In a message sent to my testing account on June 23, Google framed the adjustment as empowering me with “even more control over saved history.” The email included examples reflecting the utility of saving such media, stating, “For instance, this allows you to revisit past visual searches with Lens or continue a Search Live conversation about a song you heard.” Notably, Google did not provide analogous examples after mentioning that the saved media would contribute to AI model training; the communication simply shifted to the next point.

This change in software is significant enough to warrant careful consideration by everyday users. “It adds an additional layer of complexity that consumers must evaluate regarding their comfort with a tool they have used for a long time,” remarks Ben Winters, director of AI and privacy at the Consumer Federation of America.

I often feel overwhelmed by the constant responsibility to opt out of data training across every service I use. It leaves me feeling like a fool, worried I might overlook something hidden within these settings.

Winters perceives this development from Google as shifting the burden onto users to avoid AI training, potentially leading to widespread feelings of exhaustion and despair. “There’s a growing sense of helplessness and hopelessness for those attempting to safeguard their data, as every detail seems to be extracted,” he states.