Your uploads are being collected unless you know to turn it off
In the quiet architecture of default settings, Google has introduced a new mechanism — Search Services History — that automatically enrolls users' uploaded images, files, audio, and video into the company's AI training pipeline. Announced in June 2026 through a carefully worded email, the change reflects a broader truth of our technological moment: the data we generate in the course of daily life has become the raw material of machine intelligence, and the burden of refusal falls on the individual, not the institution. Those who wish to protect their uploads must seek out the setting, disable it, and clear what has already been gathered — a small but telling reminder that in the digital commons, silence is often taken as consent.
- Google's new Search Services History setting is active by default, meaning millions of users are already contributing their uploaded media to AI training without having made an explicit choice to do so.
- The rollout was framed as a routine reorganization of privacy controls, obscuring the significant expansion of data collection buried within the administrative language.
- The opt-out process is multi-step and non-obvious — users must navigate account settings, uncheck a specific box, confirm the change, and separately delete any media already collected.
- The scope of what is being harvested is broad: not just casual photos, but documents, receipts, voice recordings, and sensitive files that people upload for practical, everyday purposes.
- Users who had previously disabled the parent Web & App Activity setting are automatically protected, but the majority who left it enabled are now subject to the new default collection.
- Privacy advocates warn this is part of a widening pattern — as AI training demands grow, default settings across the industry are increasingly designed to capture data rather than protect it.
Google has quietly expanded its AI data collection practices with a new privacy setting called Search Services History, which automatically saves every image, file, audio clip, and video uploaded through Google Search and channels them into the company's machine learning development pipeline. The rollout began in early June with an email to users that framed the change as a routine reorganization of account controls — moving certain settings into a dedicated area within My Google Activity. The neutral, administrative tone of the message obscured what was actually happening: Google was establishing a new default that gives itself permission to use your search uploads to train its AI models.
The setting operates as a subsetting within the broader Web & App Activity control, creating a privacy hierarchy. Users who had already disabled Web & App Activity entirely are shielded automatically. But for the many who have it enabled, Search Services History is on by default — and their uploads are being collected unless they actively intervene. Opting out requires navigating to the specific setting, unchecking the Save Media option, confirming the change, and then separately reviewing and deleting any media Google has already stored.
What gives this particular weight is the nature of what people upload to Google Search. These are not always casual images — they include documents, receipts, screenshots, and voice recordings used for practical, sometimes sensitive purposes. All of it now flows into Google's AI development pipeline by default. The company frames the collection as a way to improve personalized experiences, but the deeper beneficiary is Google's expanding AI infrastructure, which it monetizes across its entire product ecosystem.
The episode is a clear illustration of how the architecture of default settings shapes behavior at scale. By making data collection the path of least resistance, Google expands its training data without requiring explicit consent each time. For users who wish to protect their information, the tools exist — but finding and using them requires awareness, initiative, and follow-through that most people will never apply.
Google quietly introduced a new way to harvest your digital life for artificial intelligence development. Last month, the company began rolling out a privacy setting called Search Services History that automatically collects every image, file, audio clip, and video you upload through Google Search—and feeds them into the machine learning systems that power everything from Google Lens to the company's latest AI models. Unless you actively disable it, your media is being saved and used to train Google's technology.
The rollout began in early June with an email notification to users. The message, titled "New privacy settings for Search services," presented the change as a routine reorganization of account controls. Google was simply moving two settings—Search Services History and Personalized Recommendations—out of the Web & App Activity section and into a dedicated area within My Google Activity. The language was neutral, almost administrative. But buried in the details was the real story: Google now has permission to take the files you search with and use them to develop and improve its AI models and safety systems.
The distinction matters because Search Services History is technically a subsetting of the larger Web & App Activity control. This means Google has created a hierarchy of privacy choices. If you've already disabled Web & App Activity entirely, you're protected—the subsetting inherits that choice automatically. But if you have Web & App Activity turned on, which many users do, the new Search Services History setting is enabled by default. Your uploads are being collected unless you know to turn it off and actually do so.
Disabling the feature requires navigating to your Search Services History settings and unchecking the box labeled "Save Media." Google will ask you to confirm the change by reading through an explanation of what you're turning off, then clicking a button to finalize it. The company also offers a separate option to view and delete any media it has already saved from your previous searches. But this is a cleanup task—you have to remember to do it, and you have to know the setting exists in the first place.
The timing of this change reflects a broader shift in how technology companies are approaching artificial intelligence. Training modern AI models requires enormous amounts of data, and companies are increasingly looking to their existing user bases as sources. Google has access to billions of searches, countless uploaded images, and vast repositories of user-generated content. By quietly making this new setting the default, the company has found a way to expand its training data without explicitly asking permission each time. Users who want to protect their uploads have to actively opt out—a burden that falls on them, not on Google.
What makes this particularly significant is the scope of what's being collected. Search uploads aren't just casual snapshots. People use Google Search to upload documents, receipts, screenshots, voice recordings, and videos for all sorts of purposes—some mundane, some sensitive. All of it now feeds into Google's AI development pipeline unless the user knows to stop it. The company frames this as a way to improve your personal experience with its products, and that's partly true. But the primary beneficiary is Google's AI capabilities, which the company can then monetize across its entire product ecosystem.
For users who care about this, the path forward is clear but requires action. Check your Search Services History settings. If Web & App Activity is enabled, go disable the Save Media option. Delete any uploads Google has already collected. And stay alert—as companies continue to find new ways to feed user data into AI systems, the default settings will likely keep working against privacy, not for it.
Citas Notables
The saved media will be used to develop and improve Google services and technologies, including AI models and safety measures— Google, in the June privacy settings notification email
La Conversación del Hearth Otra perspectiva de la historia
So Google is collecting the files people upload to search. How many people even know this is happening?
Almost none, probably. The email went out in June, but it was framed as a routine settings reorganization. Most people don't read those notifications carefully, and the real implication—that your uploads are now feeding AI training—was buried in the language.
But it's not a new practice, right? Google has always used data for AI.
True, but this is different in scale and directness. Before, your search history was one data stream among many. Now Google is explicitly collecting the actual files you upload—images, documents, audio—and saying outright that they'll use them for AI development. It's more transparent in the email, but also more aggressive in practice.
What happens if someone doesn't opt out?
Their uploads keep getting saved and fed into Google's AI models. There's no expiration date mentioned. Google will keep using that media to train its systems unless the person goes back and manually deletes it.
Is there any indication Google will change the default?
Not that I've seen. The company has structured it so Web & App Activity is the parent setting, and Search Services History is a subsetting. That design choice—making it a subsetting rather than a standalone toggle—suggests Google wanted to bury it slightly, make it less visible.
What's the actual risk here for users?
That depends on what people upload. For someone searching with screenshots of financial documents or medical records, the risk is real. Those files are now part of Google's training data. For casual photo searches, the risk is lower but still there—your images are being used to train models that Google will eventually sell access to or build products around.
So the burden is entirely on the user to know about this and act.
Completely. Google made it the default, made it a subsetting so it's less obvious, and required manual opt-out. That's the opposite of privacy-first design. It's privacy-optional design.