Open-source privacy proxy masks PII before prompts reach external AI services

Sinisa Markovic, Senior Staff Writer, Help Net Security May 1, 2026 Share Open-source privacy proxy masks PII before prompts reach external AI services Enterprise developers routinely send prompts to external large language models that contain customer emails, support transcripts, and other identifying information, often without a sanitization layer between the application and the API. Dataiku has released Kiji Privacy Proxy, an open-source local gateway that detects and masks personally identifiable information before requests leave the network. The tool sits between local applications and external AI APIs such as OpenAI and Anthropic. Inbound requests pass through a machine learning model that flags 16 or more categories of PII, including email addresses, phone numbers, Social Security numbers, credit card numbers, and IP addresses. It substitutes detected entries with realistic dummy values, forwards the masked request to the upstream API, and restores the original values in the response so the calling application receives output that matches its input. Local inference and deployment options PII detection runs on a quantized DistilBERT model executed through ONNX Runtime on the user’s machine, with no calls to outside services for the detection step itself. Latency stays under 100 milliseconds for most requests, according to the project documentation. The base model achieved a 94 percent F1 score on an industry benchmark dataset. Distribution covers three form factors. macOS users install a native Electron desktop application that configures Proxy Auto-Config so Safari and Chrome route traffic through Kiji on port 8081 without manual environment variables. Linux users run a standalone server binary and set HTTP_PROXY and HTTPS_PROXY values. A separate Chrome extension routes web requests through the proxy for users who interact with services like ChatGPT through a browser. Compliance drivers Sending PII to a third-party API can trigger obligations under GDPR, HIPAA, and CCPA, and many enterprises restrict what data may leave the corporate perimeter. A 2026 Dataiku survey of 600 CIOs found that 85 percent had seen AI projects delayed or blocked because of gaps in traceability or explainability, with privacy concerns a significant factor. Kiji Privacy Proxy is available for free on GitHub. Must read: 25 open-source cybersecurity tools that don’t care about your budget GitHub CISO on security strategy and collaborating with the open-source community Subscribe to the Help Net Security ad-free monthly newsletter to stay informed on the essential open-source cybersecurity tools. Subscribe here! More about data security Dataiku Generative AI GitHub LLMs open source OpenAI privacy software Share