Another day, another notification about a company adjusting how it handles your personal data, often with implications for AI training. Recently, Google began a global rollout of new privacy settings for its Search services, and for many users, this change means their media data could be automatically scooped up to refine Google’s artificial intelligence models.
The latest update focuses on your Search history, specifically how uploaded media – everything from photos used in reverse image searches to audio recordings from Google Translate conversations – might be retained. This data, if saved, is then leveraged to enhance the performance and capabilities of Google’s AI. It’s a significant shift in data handling that warrants your attention.
Understanding Google’s New Search Services History
This new feature, dubbed Search Services History, is nestled within Google’s account settings. What’s particularly noteworthy is that for many users, this setting, along with the crucial option to “Save media” for AI training, may already be enabled by default. This means your images, audio, and video from Search interactions could be automatically collected unless you explicitly opt out.
Google confirms that “saved media includes your images, files, and audio and video recordings from your interactions with Search services.” This encompasses a wide range of data points, such as Google Lens images, recordings from Search Live or Translate speaking practice, any content you upload, and all your voice searches. Essentially, if you interact with Google Search using anything other than text, that interaction is fair game.
The motivation behind this expanded data collection is clear: AI models thrive on diverse inputs. To make its AI smarter and more capable, Google needs more than just text data; it requires a rich diet of audio, video, and image inputs. By tapping into its vast user base across numerous services, Google gains a significant advantage in gathering this crucial training data.
A key concern arises from how long this data is kept. Google explicitly states that if your saved media is used to train their AI models, it becomes “disconnected from your Google Account” but will be retained for up to four years. This retention period applies even if you delete the original activity, meaning your data could be in Google’s AI “blender” for a substantial amount of time.
How to Review and Adjust Your Settings
It’s vital to take proactive steps to manage these new settings, ideally sooner rather than later. Once your media data is integrated into AI models, there’s little you can do to retrieve it. Google assures users that these new settings “help users get more relevant results and revisit their searches” and can be turned on or off at any time.
Here’s how you can check and adjust your Search Services History settings:
- Visit Google’s My Activity page.
- Navigate to the Search Services History tab.
- This page provides a clear overview of the data Google is saving from your Search history.
- To disable the entire setting, you can turn off the main toggle.
- Crucially, if you do not want your image uploads and other media used for AI training, ensure you uncheck the box next to “Save media.”
Making these adjustments ensures you maintain control over your personal data. The pop-up notification when turning off the “Save media” feature underscores the importance of this step, highlighting the four-year data retention policy for AI training data.
The Opt-Out Dilemma and User Exhaustion
This “opt-out” approach, where users must actively disable features rather than consciously choosing to enable them, is a recurring theme across many digital platforms. Experts like Thorin Klosowski, a senior security and privacy activist at the Electronic Frontier Foundation, argue that “opt-in” should be the bare minimum. They believe companies should be required to make a compelling case for why these features benefit users before automatically activating them.
Google’s communication about this change, often framing it as giving users “even more control,” can feel disingenuous when key aspects like AI training are downplayed. While examples are given for how saving media can be helpful (e.g., revisiting visual searches), the use for AI training is typically presented without similar illustrative benefits, creating a disconnect for users.
Ben Winters, director of AI and privacy at the Consumer Federation of America, points out that this constant need for users to perform “extra layers of math” about their comfort with data collection contributes to widespread user exhaustion. There’s an increasing feeling of powerlessness, where individuals feel that no matter how vigilant they are, some piece of their data will inevitably be collected and used.
Ultimately, while Google highlights user benefits, the onus remains squarely on the user to understand and navigate complex privacy settings. This ongoing challenge creates a sense of overwhelm, making it easy for crucial details about AI training and data retention to be overlooked in the flood of digital updates.
Source: Wired – AI
