Skip to content

PrivacyAI

Product or Programme · All products and programmes
Evidence · 1 claim (1 for) · 1 study (1 design), q2 · 0 of 1 report an effect size · 1 claim rests on one study

Description

PrivacyAI is an open-source GPT-4o-mini-based pipeline released by the authors for detecting, labeling, extracting, and verifying personally identifiable information in educational text.

Components

  • Fine-tuned GPT-4o-mini PII detection pipeline with special-identifier labeling and verifier models (code released on GitHub): A PII detection pipeline for educational text in which GPT-4o-mini is either prompted with few-shot examples or fine-tuned to surround detected entities with category-specific special identifiers (e.g., @@@Text### for student names), which are then extracted via regular expressions to ensure accurate positioning. Optional fine-tuned verifier models (with or without chain-of-thought) re-check detected entities in context to remove false positives. The article reports that "Our code is available on GitHub: https://github.com/AnonJD/PrivacyAI". (Zilyu Ji et al. (2025))

Claims

-

Key Sources

  • Zilyu Ji, Yuntian Shen, Kenneth R. Koedinger, Jionghao Lin. (2025). Enhancing the De-identification of Personally Identifiable Information in Educational Data. Journal of Educational Data Mining, Volume 17, No 2. https://github.com/AnonJD/PrivacyAI
Page metadata
Field Value
type product
id privacyai
title PrivacyAI
description PrivacyAI is an open-source GPT-4o-mini-based pipeline released by the authors for detecting, labeling, extracting, and verifying personally identifiable information in educational text.
status review
moved_from element/fine-tuned-gpt4o-mini-pii-detection-pipeline
product_kind software
generated by process:sweep-kinds on 2026-10-10
trust unverified — no one has confirmed this page's content
sources 1 — zilyu-ji-2025

Frontmatter is the page's machine-readable record — see the Schema & Guide.