mlsanityData to 14 Oct 2026

Position: Editing Large Language Models Poses Serious Safety Risks

arXiv 2502.02958 ↗First posted 5 Feb 2025cs.CLcited by 15

Not on the map. Why?

Abstract (arXiv)

Large Language Models (LLMs) contain large amounts of facts about the world. These facts can become outdated over time, which has led to the development of knowledge editing methods (KEs) that can change specific facts in LLMs with limited side effects. This position paper argues that editing LLMs poses serious safety risks that have been largely overlooked. First, we note the fact that KEs are widely available, computationally inexpensive, highly performant, and stealthy makes them an attractive tool for malicious actors. Second, we discuss malicious use cases of KEs, showing how KEs can be easily adapted for a variety of malicious purposes. Third, we highlight vulnerabilities in the AI ecosystem that allow unrestricted uploading and downloading of updated models without verification. Fourth, we argue that a lack of social and institutional awareness exacerbates this risk, and discuss the implications for different stakeholders. We call on the community to (i) research tamper-resistant models and countermeasures against malicious model editing, and (ii) actively engage in securing the AI ecosystem.

Key facts. Cited by 15 papers in the map (data through 14 Oct 2026). Builds on 59 of its 78 references that are in the map. First posted on arXiv 5 Feb 2025.

Builds on

59 of 78 references.

References · 59

  1. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 8,065
  2. 2019Parameter-Efficient Transfer Learning for NLPcited by 4,236
  3. 2021The Power of Scale for Parameter-Efficient Prompt Tuningcited by 3,795
  4. 2022Locating and Editing Factual Associations in GPTcited by 2,691
  5. 2019Language Models as Knowledge Bases?cited by 1,936
  6. 2023Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injectioncited by 1,443
  7. 2023The Rise and Potential of Large Language Model Based Agents: A Surveycited by 1,439
  8. 2022Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference timecited by 1,396
  9. 2023CAMEL: Communicative Agents for "Mind" Exploration of Large Language Model Societycited by 1,395
  10. 2021BitFit: Simple Parameter-efficient Fine-tuning for Transformer-based Masked Language-modelscited by 1,245
  11. 2017BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chaincited by 1,106
  12. 2022Mass-Editing Memory in a Transformercited by 1,001
38 more
  1. 2021Towards a Unified View of Parameter-Efficient Transfer Learningcited by 872
  2. 2020How Much Knowledge Can You Pack Into the Parameters of a Language Model?cited by 719
  3. 2021Knowledge Neurons in Pretrained Transformerscited by 601
  4. 2021Fast Model Editing at Scalecited by 498
  5. 2022Memory-Based Model Editing at Scalecited by 447
  6. 2023Safety-Tuned LLaMAs: Lessons From Improving the Safety of Large Language Models that Follow Instructionscited by 387
  7. 2020Parameter-Efficient Transfer Learning with Diff Pruningcited by 352
  8. 2020Weight Poisoning Attacks on Pre-trained Modelscited by 341
  9. 2023Can We Edit Factual Knowledge by In-Context Learning?cited by 285
  10. 2022Aging with GRACE: Lifelong Model Editing with Discrete Key-Value Adaptorscited by 271
  11. 2023Unleashing the Emergent Cognitive Synergy in Large Language Models: A Task-Solving Agent through Multi-Persona Self-Collaborationcited by 251
  12. 2023Adapted Large Language Models Can Outperform Medical Experts in Clinical Text Summarizationcited by 246
  13. 2023In-context Vectors: Making In Context Learning More Effective and Controllable Through Latent Space Steeringcited by 245
  14. 2022DyLoRA: Parameter Efficient Tuning of Pre-trained Models using Dynamic Search-Free Low-Rank Adaptationcited by 233
  15. 2024Arcee's MergeKit: A Toolkit for Merging Large Language Modelscited by 194
  16. 2023Gender bias and stereotypes in Large Language Modelscited by 187
  17. 2023PMET: Precise Model Editing in a Transformercited by 173
  18. 2024A Comprehensive Study of Knowledge Editing for Large Language Modelscited by 163
  19. 2021Composable Sparse Fine-Tuning for Cross-Lingual Transfercited by 144
  20. 2024Personalized Steering of Large Language Models: Versatile Steering Vectors Through Bi-directional Preference Optimizationcited by 118
  21. 2023EasyEdit: An Easy-to-use Knowledge Editing Framework for Large Language Modelscited by 98
  22. 2023Sparse Low-rank Adaptation of Pre-trained Language Modelscited by 96
  23. 2024BadEdit: Backdooring large language models by model editingcited by 96
  24. 2023Combating Misinformation in the Age of LLMs: Opportunities and Challengescited by 95
  25. 2024Model Editing at Scale leads to Gradual and Catastrophic Forgettingcited by 93
  26. 2024Flooding Spread of Manipulated Knowledge in LLM-Based Multi-Agent Communitiescited by 86
  27. 2024A Systematic Survey and Critical Review on Evaluating Large Language Models: Challenges, Limitations, and Recommendationscited by 85
  28. 2024Defending Large Language Models Against Jailbreak Attacks via Layer-specific Editingcited by 79
  29. 2023Conditional Adapters: Parameter-efficient Transfer Learning with Fast Inferencecited by 69
  30. 2023Residual Prompt Tuning: Improving Prompt Tuning with Residual Reparameterizationcited by 64
  31. 2023Massive Editing for Large Language Models via Meta Learningcited by 63
  32. 2024A Unified Framework for Model Editingcited by 57
  33. 2024Rebuilding ROME : Resolving Model Collapse during Sequential Model Editingcited by 50
  34. 2023An Empirical Study of Pre-Trained Model Reuse in the Hugging Face Deep Learning Model Registrycited by 47
  35. 2023Give Me the Facts! A Survey on Factual Knowledge Probing in Pre-trained Language Modelscited by 32
  36. 2023Towards Adaptive Prefix Tuning for Parameter-Efficient Language Model Fine-tuningcited by 29
  37. 2024Can Editing LLMs Inject Harm?cited by 22
  38. 2024Sowing the Wind, Reaping the Whirlwind: The Impact of Editing Language Modelscited by 20

50 most cited of 59 All 59 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 9; 2026: 6.

15 papers.

Sort:

By area:

  1. 2025Breaking Bad Tokens: Detoxification of LLMs Using Sparse Autoencoderscited by 9
  2. 2025Effective Test-Time Scaling of Discrete Diffusion through Iterative Refinementcited by 8
  3. 2025EasyEdit2: An Easy-to-use Steering Framework for Editing Large Language Modelscited by 7
  4. 2025Tracing and Reversing Edits in LLMscited by 7
  5. 2025AntiDote: Bi-level Adversarial Training for Tamper-Resistant LLMscited by 4
  6. 2026Are We Evaluating the Edit Locality of LLM Model Editing Properly?cited by 4
  7. 2025Can Fine-Tuning Erase Your Edits? On the Fragile Coexistence of Knowledge Editing and Adaptationcited by 2
  8. 2026From Backward Spreading to Forward Replay: Revisiting Target Construction in LLM Parameter Editingcited by 2
  9. 2025Detoxification of Large Language Models through Output-layer Fusion with a Calibration Modelcited by 1
  10. 2025How Robust is Model Editing after Fine-Tuning? An Empirical Study on Text-to-Image Diffusion Modelscited by 1
  11. 2026Persuasion Tokens for Editing Factual Knowledge in LLMscited by 1
  12. 2026Decoy Direction Optimization: A Post-Hoc Defense Against LLM Abliterationcited by 1
3 more
  1. 2025ADS-Edit: A Multimodal Knowledge Editing Dataset for Autonomous Driving Systemscited by 0
  2. 2026An Investigation of the NeurIPS and ICML 2025 Position Trackscited by 0
  3. 2026Selective Knowledge Edit Reversal via Gated Singular Vector Shrinkagecited by 0