Large Language Models in Misinformation Ecosystems: Misuse, Defense, and Vulnerability
Lingwei Wei, Dou Hu, Wei Zhou, Songlin Hu, Philip S. Yu
cs.CR, cs.AI, cs.SI
Submitted: 2026-07-11
Comments: 35 pages, 8 figures
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Toward a Safer Web: Multilingual Multi-Agent LLMs for Mitigating Adversarial Misinformation Attacks
- CAMOUFLAGE: Exploiting Misinformation Detection Systems Through LLM-driven Adversarial Claim Transformation
- Agentic Multi-Persona Framework for Evidence-Aware Fake News Detection
- Overcoming the Retrieval Barrier: Indirect Prompt Injection in the Wild for LLM Systems
- Uncovering Competing Poisoning Attacks in Retrieval-Augmented Generation
- TrojanRAG: Retrieval-Augmented Generation Can Be Backdoor Driver in Large Language Models
- CoAID: COVID-19 Healthcare Misinformation Dataset
- Fake News Detection After LLM Laundering: Measurement and Explanation
- UniC-RAG: Universal Knowledge Corruption Attacks to Retrieval-Augmented Generation
- MM-PoisonRAG: Disrupting Multimodal RAG with Local and Global Poisoning Attacks
- AEScorer: An Agentic Evidence-Grounded Framework for Graded Factuality Verification
- LLM-Based Adversarial Persuasion Attacks on Fact-Checking Systems
- CPA-RAG:Covert Poisoning Attacks on Retrieval-Augmented Generation in Large Language Models
- Large Language Model Agent for Fake News Detection
- FactGuard: Agentic Video Misinformation Detection via Reinforcement Learning
- RAAR: Retrieval Augmented Agentic Reasoning for Cross-Domain Misinformation Detection
- RADAR: Retrieval-Augmented Detector with Adversarial Refinement for Adaptive LLM-Generated Fake News Detection
- Adversarial Attacks on LLM-as-a-Judge Systems: Insights from Prompt Injections
- DECEIVE-AFC: Adversarial Claim Attacks against Search-Enabled LLM-based Fact-Checking Systems
- Securing Large Language Models: Addressing Bias, Misinformation, and Prompt Attacks
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