Your online data is no longer just used for ads but also for training AI. Google uses publicly available information to train its AI models, raising concerns over whether AI is even compatible with data protection laws.
People are worried companies are scraping their personal files and data without their consent or outright stealing their copyrighted materials, as highlighted by the FTC(new window). You may begin to wonder whether any of your content is safe — including your writings and other drafts in Google Docs.
Currently Google Docs says it does not scrape data for AI training purposes, but the company does scan your data and use it in various other ways. Moreover, its privacy policy(new window) doesn’t explicitly rule out the possibility of using your data for AI in the future.
This article explores how Google Docs integrates with Gemini, its large language model (LLM), and whether your data is safe from Google Docs AI scraping for training purposes. We also share how to protect your data from being used to train AI and which online document editor you can use for privacy.
- How Google Docs uses AI
- Is Google Docs scraping your data?
- Protect your data from AI training with Docs in Proton Drive
- Proton defends your online privacy
How Google Docs uses AI
Google Docs uses machine learning to offer spelling suggestions, voice typing, comment reply suggestions, and even completing documents (this last feature, Smart Compose, is only available for work or school Google accounts). You can enable or turn off these AI features by clicking Tools → Preferences in a Google Doc.
Google also has an AI assistant called Gemini that can handle all types of tasks. It’s currently working on integrating Gemini into Google Docs (if you’re a member of Google Workspace Labs, you can get early access). It goes beyond autocomplete and can be used for suggesting text, improving grammar, providing style recommendations, and even generating or clarifying content ideas.
Google’s various AI tools are clearly a big part of Google Docs’ future, which might leave you wondering what this all means for the information in your documents?
What is Google Document AI?
A quick clarification — Google also offers Google Document AI, which is an entirely different type of AI feature that allows users to quickly extract data, summarize it, and automate other tedious tasks that involve processing large amounts of documents and PDFs using artificial intelligence. Google specifically says(new window) it doesn’t use customer data to train its Document AI models.
Is Google Docs AI scraping your data?
Although Google says it does not use your Google Docs data to train its AI, Google’s privacy policy doesn’t explicitly stop the company from using your content and any other publicly available information to train their AI models. Moreover, there’s no easy way to verify that Google’s models don’t use public data since Google doesn’t disclose what data its models use(new window).
Google claims your documents and data are secure when using its AI features, stating that it doesn’t share any part of your document externally. While Google has security protocols in place, including encryption at rest and in transit, it doesn’t support end-to-end encryption, so Google itself has access to your data. Gemini and Google Docs are not open source, so there’s no way to verify the code does what Google claims, which means you’re left with no choice than trusting them to protect your privacy.
Additionally, Google Docs allows third-party extensions, which have their own privacy policies and may scrape or share your data. Be sure to check their permissions and privacy policies before granting any outside developers access to your documents.
If you want to be completely sure your data is private, the only sure solution is to deGoogle your life (stop using Google services).
Additionally, you should make sure your documents are not set to Anyone with the link can view or edit. If that’s the case then if your document becomes publicly available, the information it contains will likely be used by Google and other AI services for AI training.
If you use Gemini, you should know it stores your queries and its outputs by default. However, it does allow you to permanently delete your chats and change its settings to prevent future ones from being stored or used for AI training.
End-to-end encryption is the best way to protect your content from this. Even if your data is stored, it cannot be read or used for training AI without your explicit permission.
Protect your data from AI training with Docs in Proton Drive
Proton Drive now offers Docs, a secure and private alternative to Google Docs for creating and collaborating on online documents. Docs has all the features you expect from an online document editor, enabling you to:
- See the live cursors of your team members and their edits in real time
- Gather feedback and share ideas with comments
- Invite anyone to view or edit your docs
However, unlike Google Docs and other online document editors, we use end-to-end encryption to protect your information. With Docs, the documents you create, their version history, any notes you add, and even your keystrokes and cursor movements are all end-to-end encrypted, so they can never be fed into AI models without your permission. Like all Proton services, Docs gives you the keys to access your files so no one — not even Proton — can access them without your consent.
Get started with Proton Drive for free to encrypt your data and protect it from AI training.
Or you can use Proton to start deGoogling today.
Proton defends your online privacy
Unlike Google, Proton doesn’t profit from your data. Instead, we are exclusively supported by our paying subscribers and primarily owned by a nonprofit dedicated to promoting online privacy.
Since Proton is located in Switzerland, we adhere to some of the world’s strongest privacy laws. Furthermore, all our apps are open source and independently audited.
By signing up for a Proton account, you gain access to a secure ecosystem that protects all your data, including cloud storage, passwords, internet connection(new window), email, and calendar.
