If you use Google, you're training its AI. Here's how to opt out.

According to the recent announcement, Google has rolled out a privacy setting that lets the company retain a broader slice of user data, ranging from images and files to audio and video recordings. The move is framed as a way to boost the quality of its artificial intelligence models, but it also means everyday searches and uploads are feeding the engine unless you take action.

The expanded data collection covers media that most people associate with personal moments, photos from a vacation, voice memos, even video calls stored in Google Drive. By pulling these into its training pipelines, Google hopes to improve tasks like image recognition, speech transcription, and content recommendation across its suite of services.

From an industry perspective, the shift reflects a larger trend where tech giants leverage user‑generated content to accelerate AI development. More data generally translates into better model performance, and companies are increasingly treating everyday user activity as a massive, crowdsourced dataset. The downside is that users often hand over that data without fully understanding the long‑term implications.

Fortunately, the new UI includes an opt‑out toggle. To stop Google from harvesting your media, open your Google Account, navigate to the Data & Privacy section, locate the “Personalization and Ads” panel, and disable “Use activity to improve Google services.” You may also need to clear existing data that has already been stored for training purposes.

Privacy advocates warn that even with an opt‑out, lingering copies of previously collected material could remain in backup archives. Regulators in Europe and the United States are closely watching such practices, especially after the rise of AI‑centric legislation that emphasizes informed consent and data minimization.

You’re not alone in facing this dilemma. Other platforms, from Microsoft to Meta, are wrestling with the same balance between model improvement and user rights. The competitive pressure to ship more capable AI often pushes companies to lean on the data they already own, making transparent opt‑out mechanisms a critical differentiator.

For professionals who rely on AI tools, it’s a good habit to audit what data you allow into external models. Regularly review your privacy dashboards, prune unnecessary files, and consider storing sensitive media in encrypted vaults rather than cloud services that feed AI pipelines.

What this means for you: take a few minutes today to verify your Google settings and block further media collection, especially if you handle confidential or proprietary content. A quick workflow you can try is: ask your AI assistant to list all files in your Drive that contain personal photos, then move them to a local encrypted folder. Prompt example, “Find every image in my Google Drive taken before 2023 that includes faces, and generate a zip file for download.”