Can Prompt Anonymity Protect Your Identity From LLM Providers?

arXiv:2609.33903 · cs.CR · Submitted 2026-09-27 · Read on arXiv

cs.CR

Submitted: 2026-09-27

Updated: 2026-09-27

Code: https://github.com/dzungvpham/prompt_anonymity

Terminology

• Clustering — 21× in this paper · explained in 2 episodes
• Embeddings — 15× in this paper · explained in 2 episodes
• Benchmark — 7× in this paper · explained in 2 episodes
• Differentially Private — 5× in this paper · explained in 3 episodes
• Verification — 5× in this paper · explained in 2 episodes
• LLM-as-a-Judge — 3× in this paper · explained in 3 episodes
• Network and Distributed System Security Symposium — 3× in this paper
• Transformer — 3× in this paper · explained in 2 episodes
• Core — 2× in this paper · explained in 2 episodes
• False Acceptance Rate — 2× in this paper
• IEEE Symposium on Security and Privacy — 2× in this paper
• Large Language Models (LLMs) — 2× in this paper · explained in 6 episodes
• Theoretical Guarantees — 2× in this paper · explained in 2 episodes
• AUC — 1× in this paper · explained in 2 episodes
• Accountability — 1× in this paper · explained in 2 episodes
• Actor — 1× in this paper · explained in 3 episodes
• Agent Interactions — 1× in this paper · explained in 1 episode
• Agentic Reasoning — 1× in this paper · explained in 1 episode
• Anonymity Set — 1× in this paper · explained in 1 episode
• Benchmarking — 1× in this paper · explained in 4 episodes
• Calibration — 1× in this paper · explained in 7 episodes
• Cumulative Match Characteristics — 1× in this paper
• Distillation — 1× in this paper · explained in 2 episodes
• Generalizability — 1× in this paper · explained in 6 episodes
• LoRA Adapters — 1× in this paper · explained in 3 episodes
• Modularity — 1× in this paper · explained in 9 episodes
• Regularization — 1× in this paper · explained in 3 episodes
• Robustness — 1× in this paper · explained in 5 episodes

Sources

• EmbeddingGemma: Powerful and Lightweight Text Representations
• Agentic LLMs as Powerful Deanonymizers: Re-identification of Participants in the Anthropic Interviewer Dataset
• LLM Anonymization Against Agentic Re-Identification
• Clio: Privacy-Preserving Insights into Real-World AI Use
• Gemini Embedding 2: A Native Multimodal Embedding Model from Gemini
• On the State of the Art in Authorship Attribution and Authorship Verification
• ShareChat: A Dataset of Chatbot Conversations in the Wild
• Assessing Deanonymization Risks with Stylometry-Assisted LLM Agent

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