mlsanityData to 7 Oct 2026

The Ethical Need for Watermarks in Machine-Generated Language

arXiv 2209.03118 ↗First posted 7 Sep 2022cs.CL cs.CYcited by 30

On the map: LLM safety and security → Watermarking generated content and models

Abstract (arXiv)

Watermarks should be introduced in the natural language outputs of AI systems in order to maintain the distinction between human and machine-generated text. The ethical imperative to not blur this distinction arises from the asemantic nature of large language models and from human projections of emotional and cognitive states on machines, possibly leading to manipulation, spreading falsehoods or emotional distress. Enforcing this distinction requires unintrusive, yet easily accessible marks of the machine origin. We propose to implement a code based on equidistant letter sequences. While no such code exists in human-written texts, its appearance in machine-generated ones would prove helpful for ethical reasons.

Key facts. Cited by 30 papers in the map (data through 7 Oct 2026). Placed in Watermarking generated content and models, one of 12 directions in LLM safety and security. Builds on 10 of its 30 references that are in the map. First posted on arXiv 7 Sep 2022.

Builds on

10 of 30 references.

Closest ancestors · 1

Same direction, newest first

  1. 2021Protecting Intellectual Property of Language Generation APIs with Lexical Watermark5 Dec

Other references · 9

  1. 2020Language Models are Few-Shot Learnerscited by 32,199
  2. 2022PaLM: Scaling Language Modeling with Pathwayscited by 5,304
  3. 2022LaMDA: Language Models for Dialog Applicationscited by 1,218
  4. 2021Scaling Language Models: Methods, Analysis & Insights from Training Gophercited by 1,199
  5. 2021Ethical and social risks of harm from Language Modelscited by 957
  6. 2017Deal or No Deal? End-to-End Learning for Negotiation Dialoguescited by 202
  7. 2019DAWN: Dynamic Adversarial Watermarking of Neural Networkscited by 94
  8. 2020Imitation Attacks and Defenses for Black-box Machine Translation Systemscited by 78
  9. 2021BloomNet: A Robust Transformer based model for Bloom's Learning Outcome Classificationcited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 14; 2024: 7; 2025: 6; 2026: 3.

30 papers.

Sort:

By area:

  1. 2023A Watermark for Large Language Modelscited by 667
  2. 2023Paraphrasing evades detectors of AI-generated text, but retrieval is an effective defensecited by 401
  3. 2023DetectLLM: Leveraging Log Rank Information for Zero-Shot Detection of Machine-Generated Textcited by 207
  4. 2023Tree-Ring Watermarks: Fingerprints for Diffusion Images that are Invisible and Robustcited by 179
  5. 2023On the Reliability of Watermarks for Large Language Modelscited by 170
  6. 2023Regulating ChatGPT and other Large Generative AI Modelscited by 102
  7. 2023A Survey on Detection of LLMs-Generated Contentcited by 70
  8. 2023On the Creativity of Large Language Modelscited by 57
  9. 2023Leveraging Optimization for Adaptive Attacks on Image Watermarkscited by 50
  10. 2023Large Language Models can be Guided to Evade AI-Generated Text Detectioncited by 43
  11. 2024SoK: Watermarking for AI-Generated Contentcited by 43
  12. 2023Embarrassingly Simple Text Watermarkscited by 23
18 more
  1. 2024Mapping the Ethics of Generative AI: A Comprehensive Scoping Reviewcited by 19
  2. 2024Self-Recognition in Language Modelscited by 17
  3. 2023The Age of Synthetic Realities: Challenges and Opportunitiescited by 15
  4. 2024Optimizing Adaptive Attacks against Watermarks for Language Modelscited by 11
  5. 2024RAFT: Realistic Attacks to Fool Text Detectorscited by 8
  6. 2023Dual Use Concerns of Generative AI and Large Language Modelscited by 6
  7. 2023AI Regulation in Europe: From the AI Act to Future Regulatory Challengescited by 5
  8. 2026Tool Use Enables Undetectable Steganography in Multi-Agent LLM Systemscited by 5
  9. 2024SEFD: Semantic-Enhanced Framework for Detecting LLM-Generated Textcited by 3
  10. 2025BiMarker: Enhancing Text Watermark Detection for Large Language Models with Bipolar Watermarkscited by 3
  11. 2025Watermarking Degrades Alignment in Language Models: Analysis and Mitigationcited by 3
  12. 2025On the Detectability of LLM-Generated Text: What Exactly Is LLM-Generated Text?cited by 3
  13. 2024Ethics Whitepaper: Whitepaper on Ethical Research into Large Language Modelscited by 1
  14. 2025We Need Improved Data Curation and Attribution in AI for Scientific Discoverycited by 1
  15. 2025Temperature Matters: Enhancing Watermark Robustness Against Paraphrasing Attackscited by 0
  16. 2025Watermarking Large Language Models in Europe: Interpreting the AI Act in Light of Technologycited by 0
  17. 2026A Regulatory Placebo? The Systemic Failure of Mandatory GenAI Labelingcited by 0
  18. 2026TTMark: Pairwise Distortion-Free Watermarking Beyond Single-Token Entropycited by 0