61 articles tagged with #data-protection. AI-curated summaries with sentiment analysis and key takeaways from 50+ sources.
Researchers propose Talaria, a new confidential inference framework that protects client data privacy when using cloud-hosted Large Language Models. The system partitions LLM operations between client-controlled environments and cloud GPUs, reducing token reconstruction attacks from 97.5% to 1.34% accuracy while maintaining model performance.
Researchers have developed AloePri, the first privacy-preserving LLM inference method designed for industrial applications. The system uses collaborative obfuscation to protect input/output data while maintaining 96.5-100% accuracy and resisting state-of-the-art attacks, successfully tested on a 671B parameter model.
A study of 26 young Canadians reveals that smart voice assistants' complex privacy controls and lack of transparency discourage privacy-protective behaviors among youth. Researchers propose design improvements including unified privacy hubs, plain-language data labels, and clearer retention policies to empower young users while maintaining convenience.
A research study analyzed privacy and usability trade-offs in AI smart devices (Google Home, Alexa, Siri) used by youth, finding that Google Home scored highest for usability while Siri led in regulatory compliance. The study revealed that while youth feel capable of managing their data, technical complexity and unclear policies limit their privacy control.
OpenAI introduces new security features for ChatGPT including Lockdown Mode and Elevated Risk labels to help organizations protect against prompt injection attacks and AI-driven data exfiltration. These enterprise-focused security enhancements aim to address growing concerns about AI systems being exploited for malicious data access.
OpenAI has implemented safeguards to protect user data when AI agents interact with external links, addressing potential security vulnerabilities. The measures focus on preventing URL-based data exfiltration and prompt injection attacks that could compromise user information.
The article discusses a new differentially private framework designed to analyze AI chatbot usage patterns while protecting user privacy. This approach allows researchers to gain valuable insights into how users interact with AI systems without compromising individual data security.
The article discusses methods for running privacy-preserving machine learning inferences on Hugging Face endpoints. This technology allows users to perform AI model computations while protecting sensitive input data from being exposed to the service provider.
Hugging Face has partnered with Wiz Research to enhance AI security measures. This collaboration aims to improve security protocols and protect AI models and datasets on the Hugging Face platform.
Hugging Face has been selected to participate in the French Data Protection Agency's (CNIL) enhanced support program. This program provides regulatory guidance and support to help companies navigate data protection compliance requirements in France.
The article reviews nine privacy-focused AI tools as alternatives to Big Tech AI platforms that extensively collect user data. It evaluates different AI tools based on various threat models to help users choose options that better protect their privacy.
