# Model-Agnostic PII Detector for LLMs

Digest AI · Generative AI & Models · published 2026-09-10T16:02:16Z

Canonical: https://digestai.news/story/model-agnostic-pii-detector-for-llms

## Summary

A new model-agnostic detector for personally identifiable information (PII) has been released, designed to run on any large language model (LLM) managed on Amazon Bedrock. The detector, evaluated on five public PII corpora across nine LLM-based detectors, including the OpenAI PrivacyFilter, can target new entities by editing prompts and is agnostic to the backend. The package, available in the sample-llm-pii-detection repository, is configurable and can be run on your own data. The detector is built from four parts: a prompt defining the schema, a backend running the model, a parsing-and-offset layer turning the response into located spans, and a thin call sequence tying them together. The detector is evaluated against eight other LLM-based detectors, including the OpenAI PrivacyFilter, and is scored on a common ground truth. The package ships the Amazon Bedrock adapter and can run against a managed model on Amazon Bedrock or an open model hosted on Amazon Elastic Compute Cloud (Amazon EC2).

## Key points

- Model-agnostic PII detector for LLMs
- Runs on any LLM managed on Amazon Bedrock
- Can target new entities by editing prompts

## Why it matters

This model-agnostic PII detector can be applied to any LLM, making it a versatile tool for detecting personally identifiable information.

## Sources

1. [Model-agnostic PII detection with LLMs](https://aws.amazon.com/blogs/machine-learning/model-agnostic-pii-detection-with-llms) (AWS Machine Learning Blog, 2026-09-10, primary source)

## Cite

Digest AI, "Model-Agnostic PII Detector for LLMs", 10 September 2026, https://digestai.news/story/model-agnostic-pii-detector-for-llms

---

Written by Digest AI's editorial model from the linked sources; the sources are the record. Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse
JSON: https://digestai.news/story/model-agnostic-pii-detector-for-llms.json