Proton Launches Encrypted Docs with Commitment to User Privacy, But Conversion Issues Persist
Key features of Proton Docs: Proton, a privacy-focused Swiss tech firm, has launched Proton Docs, an encrypted document editing and sharing feature within its Proton Drive product:
- Documents are end-to-end encrypted, preventing their use in training generative AI models unless users actively choose to feed their content to an AI.
- Users can share documents with others by adding their email addresses and revoking permissions later if needed.
- The new feature aims to make Proton Drive a more viable competitor to Google Drive by enabling real-time editing and collaboration.
Conversion issues and limitations: Early tests reveal that Proton Docs still has some kinks to work out in converting and displaying existing formatted documents:
- An 80-page .docx file lost formatting, text, headers, footers, and page breaks when opened in Proton Docs, despite displaying accurately in the Drive preview.
- Shorter .docx files fared better but still saw some formatting like strikethroughs and heading styles stripped out.
- .txt files seemed to convert with the fewest issues, while .pages files are not supported at all.
- Proton acknowledged these issues and says it is working on fixes and plans to expand supported file types in the future.
Proton’s privacy stance: Unlike Google, which is integrating its Gemini AI across its Drive suite and using it to analyze user data, Proton has taken a strong public stance against using customer files to train AI models:
- Proton indexes Drive files locally on user devices to enable searching without sending data to the cloud.
- The company states users “should be free to work on drafts without your data being collected, scanned, and abused for profit.”
- This commitment to privacy sets Proton apart but may also slow down its ability to match the AI-powered features of competitors.
Looking ahead: While Proton Docs is not yet a full-fledged Google Docs replacement, the company’s user-first privacy stance is a compelling differentiator as concerns grow over how tech giants are using personal files to train AI. However, Proton will need to quickly address conversion issues and expand functionality to become a viable alternative for privacy-conscious users. Its measured approach reflects the difficult balance between competing on features and staying true to its core values around user control and consent.
Proton Launches Encrypted Docs, Promises No AI Training