DocCloak – Browser-Based Document Anonymizer for AI Submissions (github.com)

DocCloak has announced a new open-source, browser-based document anonymizer specifically designed for AI submissions, which ensures that personally identifiable information (PII) is securely stripped from documents before they are processed by AI systems. Operating entirely within the user's browser, DocCloak eliminates the risks associated with data leaks by not sending any information to external servers or APIs. It enables professionals—ranging from lawyers to healthcare providers—to anonymize sensitive data seamlessly, securely utilizing local machine learning models to detect and redact various types of PII, including names, emails, and more. The significance of DocCloak lies in its innovative approach to privacy-conscious AI interaction. By ensuring that AI tools can be leveraged without compromising client confidentiality or sensitive information, DocCloak addresses a critical need in industries where data privacy is paramount. The software supports various document formats (.doc, .pdf, image files) and provides features for easy restoration of redacted information in AI responses. It emphasizes user control and security with a consent-first setup, verification of data integrity, and offline functionality, making it an essential tool for anyone handling sensitive data in AI applications.