Pii
PulseAugur coverage of Pii — every cluster mentioning Pii across labs, papers, and developer communities, ranked by signal.
- instance of Mohamed Elmahallawy 90%
- used by General Data Protection Regulation 70%
- used by Health Insurance Portability and Accountability Act 70%
- used by named-entity recognition 70%
- used by Presidio 70%
- used by Gotit.pub 60%
- used by alphaXiv 60%
- uses ScienceCast 60%
- uses CatalyzeX 50%
- used by DagsHub 50%
8 day(s) with sentiment data
-
New research compares PII redaction methods for LLM pipelines
A new research paper compares four methods for redacting personally identifiable information (PII) before sending text to large language models (LLMs). The study evaluated pure regular expressions (regex), Microsoft Pre…
-
New framework ReCoPII improves PII masking dataset creation
Researchers have developed ReCoPII, a novel framework for creating contextual Personally Identifiable Information (PII) masking datasets. This framework addresses limitations of existing methods by planning entity relat…
-
reskSecure introduces LLM firewall to prevent PII exfiltration
A new security tool called reskSecure has been introduced to protect customer-support AI assistants from sensitive data exfiltration. The tool acts as an LLM firewall, blocking forbidden phrases and unauthorized tool ca…
-
New research reveals type-dependent privacy risks in code LLMs
A new research paper published on arXiv explores the privacy risks associated with large language models for code (LLM4Code). The study, authored by Hua Yang, investigates how different types of personally identifiable …
-
LLM PII Handling: Tokenization vs. Masking for GDPR Compliance
A new approach to handling Personally Identifiable Information (PII) in Large Language Model (LLM) prompts involves reversible tokenization, which replaces sensitive data with stand-in tokens before a request is sent to…
-
PgSafe MCP secures LLM access to PostgreSQL databases
A new open-source tool called PgSafe MCP has been developed to enhance the security of connecting large language models like Claude Desktop and Cursor to PostgreSQL databases. This tool addresses three primary risks: de…
-
AI shifts from task automation to workflow orchestration, demanding robust guardrails
The articles discuss the evolving landscape of AI, distinguishing between task automation and workflow orchestration. Agentic AI is presented as a tool for managing complex workflows by evaluating outcomes and deciding …
-
New framework protects PII in LLM interactions via reversible pseudo-anonymization
A new framework called Veil-then-Reveal (VtR) has been proposed to protect Personally Identifiable Information (PII) during interactions with large language models (LLMs). This user-side approach aims to prevent risks a…
-
New PII detector achieves high recall with lower cost · arXiv research
Researchers have developed a new PII detection model that is both broad and efficient, addressing the limitations of existing detectors which often rely on complex architectures and limited synthetic datasets. This new …
-
New JEPA-Prune method removes PII from fine-tuned LLMs with minimal utility loss
Researchers have developed a new method called JEPA-Prune to address privacy concerns in large language models (LLMs) after fine-tuning. This technique aims to remove personally identifiable information (PII) without si…
-
New PrivacyGuard system significantly reduces PII leakage in cloud GUI agents
A new research paper introduces PrivacyGuard, a system designed to protect personally identifiable information (PII) when cloud GUI agents interact with remote planners. PrivacyGuard employs source-linked sanitization a…
-
LLM Integration Guide: Versioning, Monitoring, and Security for SaaS
This article provides practical advice for integrating large language models (LLMs) into production SaaS applications. It emphasizes treating prompts as code, versioning them in Git, and implementing strategies for batc…
-
LLM Observability: Tracking Key Signals for Safe AI Deployment
Implementing observability for Large Language Models (LLMs) is crucial for safe and scalable AI application development. This involves integrating tools like OpenTelemetry to track key signals such as latency, token usa…
-
MuleSoft agentic flow testing strategy detailed across five layers
This article details a comprehensive testing strategy for an agentic change-approval system built on MuleSoft. It outlines a five-layer testing approach, starting with deterministic MUnit tests for individual tools and …
-
New Firewall Tool Protects Customer Support LLMs from PII Leaks
A new open-source tool called Resk has been developed to act as a firewall for customer-support AI assistants, preventing the exfiltration of Personally Identifiable Information (PII). The tool, built with FastAPI and R…
-
New Locket Framework Enhances LLM Privacy and Access Control
Researchers have developed a new framework called Locket that enhances security and privacy for large language models (LLMs). Locket uses lightweight adapters, each associated with a specific access policy, and a gating…
-
Hugging Face paper reveals reference choice drastically alters benchmark scores
A new paper from Hugging Face highlights a significant issue in multilingual benchmark scoring, demonstrating that the choice of reference data can drastically alter benchmark results. Researchers found that by simply s…
-
AI Chatbot Guardrail Flaw: PII Leak Exposed by Incomplete Monitoring
A developer discovered a critical security flaw in their AI chatbot's guardrail system where sensitive customer data was being leaked. The system incorrectly reported that a Personally Identifiable Information (PII) att…
-
AWS launches serverless PII redaction pipeline with Amazon Bedrock
AWS has introduced a new serverless pipeline for Personally Identifiable Information (PII) redaction using Amazon Bedrock Data Automation (BDA). This solution aims to automate the detection and removal of sensitive data…
-
New RAG-CT defense mitigates privacy risks in LLM generation systems
Researchers have developed a new defense mechanism called RAG-CT to address privacy risks in Retrieval-Augmented Generation (RAG) systems. These systems, which enhance Large Language Models (LLMs) by grounding responses…