# Unique Security and Privacy Threats of Large Language Models: A Comprehensive Survey

arXiv: https://arxiv.org/abs/2406.07973 · HTML: https://mlsanity.com/paper/2406.07973/

## Key facts

- Cited by 23 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 12 Jun 2024.

## Built on by (most cited first, 23 of 23)

- [The Emerged Security and Privacy of LLM Agent: A Survey with Case Studies](https://mlsanity.com/paper/2407.19354/) (2024)
- [When Machine Unlearning Meets Retrieval-Augmented Generation (RAG): Keep Secret or Forget Knowledge?](https://mlsanity.com/paper/2410.15267/) (2024)
- [1-2-3 Check: Enhancing Contextual Privacy in LLM via Multi-Agent Reasoning](https://mlsanity.com/paper/2508.07667/) (2025)
- [The Trust Paradox in LLM-Based Multi-Agent Systems: When Collaboration Becomes a Security Vulnerability](https://mlsanity.com/paper/2510.18563/) (2025)
- [SoK: Trust-Authorization Mismatch in LLM Agent Interactions](https://mlsanity.com/paper/2512.06914/) (2025)
- [VortexPIA: Indirect Prompt Injection Attack against LLMs for Efficient Extraction of User Privacy](https://mlsanity.com/paper/2510.04261/) (2025)
- [SoK: Privacy Risks and Mitigations in Retrieval-Augmented Generation Systems](https://mlsanity.com/paper/2601.03979/) (2026)
- [PromptExp: Multi-granularity Prompt Explanation of Large Language Models](https://mlsanity.com/paper/2410.13073/) (2024)
- [Toward a Safe Internet of Agents](https://mlsanity.com/paper/2512.00520/) (2025)
- [CAPID: Context-Aware PII Detection for Question-Answering Systems](https://mlsanity.com/paper/2602.10074/) (2026)
- [Guided Collaboration in Heterogeneous LLM-Based Multi-Agent Systems via Entropy-Based Understanding Assessment and Experience Retrieval](https://mlsanity.com/paper/2602.13639/) (2026)
- [Triaging Threats to Specialized Guardrails](https://mlsanity.com/paper/2605.30693/) (2026)
- [FORGE: Research-Trajectory Hijacking Attacks on Deep Research Agents](https://mlsanity.com/paper/2607.04718/) (2026)
- [Hidden Risks of Jev: An Empirical Study of Security, Privacy, and Dual Use](https://mlsanity.com/paper/2610.04985/) (2026)
- [Routing-Based Continual Learning for Multimodal Large Language Models](https://mlsanity.com/paper/2511.01831/) (2025)
- [Towards Privacy-Preserving Mental Health Support with Large Language Models](https://mlsanity.com/paper/2601.01993/) (2026)
- [A Case Study on the Impact of Anonymization Along the RAG Pipeline](https://mlsanity.com/paper/2604.15958/) (2026)
- [AVISE: Framework for Evaluating the Security of AI Systems](https://mlsanity.com/paper/2604.20833/) (2026)
- [On the Privacy of LLMs: An Ablation Study](https://mlsanity.com/paper/2605.02255/) (2026)
- [On Understanding, Identifying, and Mitigating Vulnerabilities in Agentic Large Language Models](https://mlsanity.com/paper/2608.10530/) (2026)
- [Where A Small Language Model Helps in Invoice Categorisation, Understood Through Embedding Geometry](https://mlsanity.com/paper/2608.18033/) (2026)
- [Privacy in Personalized AI Is a System Property, Not Just a Model Property](https://mlsanity.com/paper/2609.38289/) (2026)
- [Forgetting is Not Always Bad: A Neuro-Inspired Memory Repair Mechanism for Poisoned LLM Agents](https://mlsanity.com/paper/2610.13267/) (2026)

## Abstract (arXiv)

With the rapid development of artificial intelligence, large language models (LLMs) have made remarkable advancements in natural language processing. These models are trained on vast datasets to exhibit powerful language understanding and generation capabilities across various applications, including chatbots, and agents. However, LLMs have revealed a variety of privacy and security issues throughout their life cycle, drawing significant academic and industrial attention. Moreover, the risks faced by LLMs differ significantly from those encountered by traditional language models. Given that current surveys lack a clear taxonomy of unique threat models across diverse scenarios, we emphasize the unique privacy and security threats associated with four specific scenarios: pre-training, fine-tuning, deployment, and LLM-based agents. Addressing the characteristics of each risk, this survey outlines and analyzes potential countermeasures. Research on attack and defense situations can offer feasible research directions, enabling more areas to benefit from LLMs.