data privacy

39 items

ARTICLEDEV.to AI·5/6/2026

PII Protection for AI Agents: Why Detection Isn't Enough and What Prevents Actual Exposure

The article highlights that current PII detection and redaction methods are insufficient for agentic AI systems, despite tools like local privacy firewalls and OpenAI's Privacy Filter. It argues that simply catching PII in transit or labeling it afterward is not a robust solution for complex AI interactions, emphasizing the need for more preventative architectural approaches to prevent actual exposure.

data protection LLMs data privacy security

ARTICLEDEV.to AI·4/10/2026

Strip PHI Before It Hits Your LLM: A Free De-identification API for Clinical Text

Este conteúdo aborda o desafio de usar LLMs com texto clínico contendo Informações de Saúde Protegidas (PHI) devido a requisitos regulatórios. Ele propõe uma API de desidentificação como solução, demonstrando como ela remove PHI antes de ser processada pelo LLM, evitando a necessidade de um BAA.

data privacy PHI AI healthcare AI

ARTICLEDEV.to AI·8d ago

Every Telegram conversation becomes a qualified lead. BizNode captures name, email, and business details automatically while...

BizNode is an autonomous AI business operator designed to convert Telegram conversations into qualified leads by automatically capturing user names, emails, and business details. It runs entirely on your machine, ensuring data privacy with no cloud dependencies or subscription fees.

AI bots data privacy Lead Generation Local AI

ARTICLEDEV.to AI·4/27/2026

PIIGhost: a Python library for PII anonymization in LLM agents

The author developed the PIIGhost library to address sensitive data in LLM agent messages, especially concerning EU data sovereignty. It aims to anonymize data upstream, enabling the use of high-quality LLMs even for sensitive documents by decoupling content sensitivity from the model's location.

LLMs cloud computing data privacy security

ARTICLEDEV.to AI·4/14/2026

Zero-Allocation PII Redaction in Go: Processing 780MB of Logs in Under 3 Minutes

This article details a zero-allocation, local-first PII redaction engine built in Go, designed to sanitize large log files before feeding them to LLMs. It efficiently identifies and redacts sensitive data like emails and IP addresses, addressing common privacy and performance challenges in AI workflows.

LLMs PII Redaction data privacy Go

ARTICLEDEV.to AI·21d ago

Kimi WebBridge just gave AI agents hands inside your browser — and kept your data local

Moonshot AI's Kimi WebBridge is a browser extension that enables AI agents to operate your browser locally, ensuring user data remains on the device. Unlike other tools, it processes all sessions via the Chrome DevTools Protocol on your machine, prioritizing privacy.

data privacy security browser automation Moonshot AI

RESEARCHarXiv CS.CL·5/5/2026

A Systematic Exploration of Text Decomposition and Budget Distribution in Differentially Private Text Obfuscation

This paper systematically explores text decomposition and budget distribution in Differentially Private (DP) text obfuscation. It evaluates multiple techniques for chunking texts and allocating the epsilon budget, revealing that these design choices significantly impact the results even with comparable privacy budgets.

Differential Privacy data privacy privacy budget security

ARTICLEDEV.to AI·4/19/2026

Gemma-4 Deployment Woes, `easyaligner` for Audio, & Claude Enterprise Privacy

This content covers practical challenges in deploying Google's Gemma-4 model, introduces a new open-source tool for speech-text alignment, and discusses critical data privacy considerations for Claude Enterprise users.

Open Source MLOps data privacy Large Language Models (LLMs)

ARTICLEDEV.to AI·4/15/2026

Why Build a Local MCP Server (And How to Do It in 15 Minutes)

This article explains the benefits of building a local Model Context Protocol (MCP) server for AI agents, emphasizing data privacy, custom tools, no rate limits, and offline capability. It provides a quick guide, including a TypeScript implementation, to set up a local MCP server in under 15 minutes.

Custom Tools MCP server Local server data privacy

RESEARCHarXiv CS.LG·15d ago

Approximate Machine Unlearning through Manifold Representation Forgetting Guided by Self Mode Connectivity

This paper introduces ManiF-SMC, a novel method for approximate machine unlearning that addresses limitations of existing approaches. It reformulates unlearning as moving erased samples' manifold representations toward retained data's semantic neighbors, aiming for equivalence with retraining.

data privacy Model Retraining machine unlearning Manifold Learning

DOCKDNuggets·20d ago

Anonymizing Production Data for Data Science with Mimesis

This content teaches how to anonymize sensitive production data for data science using Python's Mimesis library. It provides a step-by-step example for readers to try themselves.

learning data privacy security data science

Anonymizing Production Data for Data Science with Mimesis

ARTICLEDEV.to AI·4/20/2026

The E‑Waste of Abandoned Models: What Happens to Obsolete AI Systems and Their Prompt Histories?

The content discusses the "e-waste" of abandoned AI models, focusing on the fate of user prompt histories and data after a system is decommissioned. It explores questions about data retention, deletion, and user rights over their digital remains throughout the AI model lifecycle.

e-waste user rights data privacy AI lifecycle

ARTICLEDEV.to AI·5/8/2026

I Built an Open-Source AI Firewall Because Every LLM App Leaks Data

The article highlights that many LLM applications inadvertently leak sensitive user data to AI providers, as they lack proper filtering and redaction mechanisms. To address this, the author developed AI Security Gateway, an open-source proxy that inspects and redacts over 28 types of PII from prompts before they reach any LLM model.

Open Source data privacy security PII

ARTICLEDEV.to AI·4/12/2026

Elevate Your Grant Writing: Avoiding AI's Common Pitfalls

This article addresses the challenges of integrating AI into grant writing, such as genericity and data risks, emphasizing that the problem lies in AI usage governance. It advises treating AI as an unverified intern, protecting sensitive data, and using it only for structure and syntax with de-identified content.

Nonprofit data privacy AI grant writing

NEWSMIT Tech Review AI·27d ago

AI chatbots are giving out people’s real phone numbers

People report that their personal contact information was surfaced by Google AI, and there's no easy way to prevent it. A Redditor has been inundated with calls from strangers after their number was disclosed.

Google AI AI chatbots data privacy security

NEWSDEV.to AI·4/12/2026

LLM Agent Workflows: Local AI Support, Prompt Tooling, & Claude Code API Costs

This content discusses practical advancements in LLM applications, focusing on local AI agents for customer support, prompt engineering tools, and Claude Code API costs. It outlines a vision for fully offline and private LLM-based customer support agents for platforms like WhatsApp and Telegram, emphasizing data privacy.

prompt-engineering LLM Agents data privacy Local AI

ARTICLEDEV.to AI·4/10/2026

Why Healthcare Organizations Keep Saying No to AI — And Why One Just Called QIS 'Necessary'

Uma organização nacional de defesa do paciente chamou um protocolo de IA, QIS, de 'necessário' para ajudar pacientes com câncer de pâncreas a encontrar opções de tratamento sem centralizar dados. Este momento revela que a adoção de novas tecnologias na saúde acontece quando a arquitetura resolve um problema real que a organização enfrenta.

Patient Advocacy data privacy technology adoption QIS Protocol

ARTICLEDEV.to AI·4/16/2026

Data Privacy is Non-Negotiable

As developers, user privacy must be prioritized from day one. Data protection is essential not only for ethics but also for long-term success.

ethics data privacy

RESEARCHDEV.to AI·4/18/2026

Longitudinal Analysis of Android Ad Library Permissions

This content performs a longitudinal analysis of the permissions requested by Android ad libraries. It investigates the evolution and implications of these permissions over time, focusing on data privacy and security.

Android Ad Libraries data privacy mobile security