PrivacyScrubber Launches Zero-Trust MCP Server to Redact PII in AI IDEsPrivacyScrubber introduces a local-only Model Context Protocol (MCP) server, allowing developers to scrub PII from AI prompts directly inside Cursor & Windsurf.
By: PrivacyScrubber As enterprise adoption of AI coding assistants accelerates, software developers frequently copy-paste database schemas, customer records, and system logs containing sensitive PII into AI interfaces for debugging. Under regulations like GDPR, HIPAA, and SOC 2, transmitting this data to public cloud models constitutes a severe compliance violation and a leak of corporate intellectual property. PrivacyScrubber addresses this vulnerability via its Zero-Trust Data Sanitization (ZTDS) architecture. Running entirely on the user's local machine, the new MCP server intercepts prompts inside the editor, replacing sensitive emails, names, phone numbers, and keys with secure token placeholders (e.g., [EMAIL_1], [NAME_2]). After the AI generates a response, the developer can run a local "reverse scrub" to restore the original values. Key Features of the PrivacyScrubber Launch: 1. 100% Local Processing (Airplane Mode Verified): All redaction, regex profiling, and text sanitization occur in local RAM. No data is stored, logged, or sent to external servers, satisfying CISO and security review requirements. 2. Native IDE Integration: 3. Reversible Tokenization: "Our MCP server bridges the gap between high-security compliance and frictionless developer workflows," said Ilya Sibiryakov, Founder of PrivacyScrubber. "Developers should not have to choose between writing secure code and using AI assistants. By keeping all PII redaction local, we offer a Zero-Trust solution that runs entirely at the edge." Availability and Pricing: PrivacyScrubber's MCP server is open-source and available on GitHub under the MIT license at https://github.com/ End
|
|