CLIOPATRA: Extracting Private Information from LLM Insights

arXiv:2603.09781 · cs.CR · Submitted 2026-03-10 · Read on arXiv

cs.CR

Submitted: 2026-03-10

Updated: 2026-10-13

Code: https://github.com/microsoft/presidio

Project page: https://microsoft.github.io/presidio/faq

Terminology

• Aggregation — 1× in this paper · explained in 1 episode
• Clustering — 1× in this paper · explained in 2 episodes

Sources

• Is My Data in Your Retrieval Database? Membership Inference Attacks Against Retrieval Augmented Generation
• Design Patterns for Securing LLM Agents against Prompt Injections
• Phantom: General Backdoor Attacks on Retrieval Augmented Language Generation
• Meta-SecAlign: Training LLMs against Prompt Injection for Robust Agents
• Defeating Prompt Injections by Design
• The Llama 3 Herd of Models
• Which Economic Tasks are Performed with AI? Evidence from Millions of Claude Conversations
• Defending Against Indirect Prompt Injection Attacks With Spotlighting
• Gemma 3 Technical Report
• The Attacker Moves Second: Stronger Adaptive Attacks Bypass Defenses Against Llm Jailbreaks and Prompt Injections
• Clio: Privacy-Preserving Insights into Real-World AI Use
• Qwen3 Technical Report

Related papers

• SoK: AI-Augmented Binary Reversing
• Relaxed Sender Anonymity for CBDC Interbank Settlement: A Zero-Knowledge Approach on Permissioned EVM
• Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages
• Efficient Fuzzy PSI under One-Sided Assumptions
• Sealing the Audit-Runtime Gap for LLM Skills
• Token Composition: A Graph Based on EVM Logs