mlsanityData to 14 Oct 2026

Clio: Privacy-Preserving Insights into Real-World AI Use

arXiv 2412.13678 ↗First posted 18 Dec 2024cs.CY cs.AI cs.CLcited by 72

Not on the map. Why?

Abstract (arXiv)

How are AI assistants being used in the real world? While model providers in theory have a window into this impact via their users' data, both privacy concerns and practical challenges have made analyzing this data difficult. To address these issues, we present Clio (Claude insights and observations), a privacy-preserving platform that uses AI assistants themselves to analyze and surface aggregated usage patterns across millions of conversations, without the need for human reviewers to read raw conversations. We validate this can be done with a high degree of accuracy and privacy by conducting extensive evaluations. We demonstrate Clio's usefulness in two broad ways. First, we share insights about how models are being used in the real world from one million Claude.ai Free and Pro conversations, ranging from providing advice on hairstyles to providing guidance on Git operations and concepts. We also identify the most common high-level use cases on Claude.ai (coding, writing, and research tasks) as well as patterns that differ across languages (e.g., conversations in Japanese discuss elder care and aging populations at higher-than-typical rates). Second, we use Clio to make our systems safer by identifying coordinated attempts to abuse our systems, monitoring for unknown unknowns during critical periods like launches of new capabilities or major world events, and improving our existing monitoring systems. We also discuss the limitations of our approach, as well as risks and ethical concerns. By enabling analysis of real-world AI usage, Clio provides a scalable platform for empirically grounded AI safety and governance.

Key facts. Cited by 72 papers in the map (data through 14 Oct 2026). Builds on 27 of its 104 references that are in the map. First posted on arXiv 18 Dec 2024.

Builds on

27 of 104 references.

References · 27

  1. 2019Sentence-BERT: Sentence Embeddings using Siamese BERT-Networkscited by 9,170
  2. 2016Communication-Efficient Learning of Deep Networks from Decentralized Datacited by 7,669
  3. 2018UMAP: Uniform Manifold Approximation and Projection for Dimension Reductioncited by 2,898
  4. 2021Ethical and social risks of harm from Language Modelscited by 966
  5. 2022Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learnedcited by 729
  6. 2024WildChat: 1M ChatGPT Interaction Logs in the Wildcited by 629
  7. 2023LMSYS-Chat-1M: A Large-Scale Real-World LLM Conversation Datasetcited by 461
  8. 2023A Survey on Large Language Model (LLM) Security and Privacy: The Good, the Bad, and the Uglycited by 387
  9. 2021Understanding Dataset Difficulty with V-Usable Informationcited by 291
  10. 2022What Does it Mean for a Language Model to Preserve Privacy?cited by 195
  11. 2024Security and Privacy Challenges of Large Language Models: A Surveycited by 186
  12. 2023Sociotechnical Safety Evaluation of Generative AI Systemscited by 157
15 more
  1. 2023Harms from Increasingly Agentic Algorithmic Systemscited by 135
  2. 2019The Ethics of AI Ethics -- An Evaluation of Guidelinescited by 129
  3. 2024The Ethics of Advanced AI Assistantscited by 107
  4. 2022The Text Anonymization Benchmark (TAB): A Dedicated Corpus and Evaluation Framework for Text Anonymizationcited by 95
  5. 2024Concept Induction: Analyzing Unstructured Text with High-Level Concepts Using LLooMcited by 78
  6. 2020Privacy in Deep Learning: A Surveycited by 71
  7. 2020Differentially Private Representation for NLP: Formal Guarantee and An Empirical Study on Privacy and Fairnesscited by 67
  8. 2023Privacy Issues in Large Language Models: A Surveycited by 60
  9. 2023Goal-Driven Explainable Clustering via Language Descriptionscited by 58
  10. 2024Trust No Bot: Discovering Personal Disclosures in Human-LLM Conversations in the Wildcited by 47
  11. 2023The Shifted and The Overlooked: A Task-oriented Investigation of User-GPT Interactionscited by 45
  12. 2024TnT-LLM: Text Mining at Scale with Large Language Modelscited by 26
  13. 2024First-Person Fairness in Chatbotscited by 20
  14. 2024The Use of Generative Search Engines for Knowledge Work and Complex Taskscited by 18
  15. 2024The Role of Governments in Increasing Interconnected Post-Deployment Monitoring of AIcited by 7

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 30; 2026: 42.

72 papers.

Sort:

By area:

  1. 2025GDPval: Evaluating AI Model Performance on Real-World Economically Valuable Taskscited by 158
  2. 2025Organize the Web: Constructing Domains Enhances Pre-Training Data Curationcited by 76
  3. 2025Toward an Evaluation Science for Generative AI Systemscited by 45
  4. 2025The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating Itcited by 45
  5. 2025EvalTree: Profiling Language Model Weaknesses via Hierarchical Capability Treescited by 27
  6. 2026How AI Impacts Skill Formationcited by 24
  7. 2025CUPID: Evaluating Personalized and Contextualized Alignment of LLMs from Interactionscited by 22
  8. 2025HiBayES: A Hierarchical Bayesian Modeling Framework for AI Evaluation Statisticscited by 18
  9. 2026Detecting Safety Violations Across Many Agent Tracescited by 14
  10. 2025Generative Value Conflicts Reveal LLM Prioritiescited by 13
  11. 2025The Lock-in Hypothesis: Stagnation by Algorithmcited by 11
  12. 2025Culture is Everywhere: A Call for Intentionally Cultural Evaluationcited by 10
38 more
  1. 2026Align Once, Benefit Multilingually: Enforcing Multilingual Consistency for LLM Safety Alignmentcited by 9
  2. 2025Benchmarking Misuse Mitigation Against Covert Adversariescited by 8
  3. 2025The Pluralistic Moral Gap: Understanding Judgment and Value Differences between Humans and Large Language Modelscited by 7
  4. 2026DiscoverLLM: From Executing Intents to Discovering Themcited by 6
  5. 2025Real-World Gaps in AI Governance Researchcited by 5
  6. 2025LLMs unlock new paths to monetizing exploitscited by 5
  7. 2025Can Large Language Models Really Recognize Your Name?cited by 4
  8. 2025Urania: Differentially Private Insights into AI Usecited by 4
  9. 2025When Models Know More Than They Can Explain: Quantifying Knowledge Transfer in Human-AI Collaborationcited by 4
  10. 2025Attention to Non-Adopterscited by 4
  11. 2026Taxonomy of the Retrieval System Framework: Pitfalls and Paradigmscited by 4
  12. 2026Automatically Finding Reward Model Biasescited by 4
  13. 2026Implicit Turn-Wise Policy Optimization for Proactive User-LLM Interactioncited by 4
  14. 2025Dataset Featurization: Uncovering Natural Language Features through Unsupervised Data Reconstructioncited by 3
  15. 2025Societal Impacts Research Requires Benchmarks for Creative Composition Taskscited by 3
  16. 2025HERCULES: Hierarchical Embedding-based Recursive Clustering Using LLMs for Efficient Summarizationcited by 3
  17. 2025Show or Tell? Modeling the evolution of request-making in Human-LLM conversationscited by 3
  18. 2025Evalet: Evaluating Large Language Models through Functional Fragmentationcited by 3
  19. 2025Breaking Down Bias: On The Limits of Generalizable Pruning Strategiescited by 2
  20. 2025A closer look at how large language models trust humans: patterns and biasescited by 2
  21. 2025Who's Asking? Simulating Role-Based Questions for Conversational AI Evaluationcited by 2
  22. 2025When Facts Change: Probing LLMs on Evolving Knowledge with evolveQAcited by 2
  23. 2025A Hierarchical Imprecise Probability Approach to Reliability Assessment of Large Language Modelscited by 2
  24. 2026Who Defines "Best"? Towards Interactive, User-Defined Evaluation of LLM Leaderboardscited by 2
  25. 2026Reflections and New Directions for Human-Centered Large Language Modelscited by 2
  26. 2026Deep Reasoning in General Purpose Agents via Structured Meta-Cognitioncited by 2
  27. 2026SynAE: A Framework for Measuring the Quality of Synthetic Data for Tool-Calling Agent Evaluationscited by 2
  28. 2026AI Fiction in the Wildcited by 2
  29. 2026Google's AI & Economy ATLAS v1.0: Mapping Gemini Usage in the Economycited by 2
  30. 2026Are LLMs becoming similarly creative? Evidence from three years of modelscited by 2
  31. 2026What LLMs Think When You Don't Tell Them What to Think About?cited by 1
  32. 2026Abstractive Red-Teaming of Language Model Charactercited by 1
  33. 2026BenchBrowser: Retrieving Evidence for Evaluating Benchmark Validitycited by 1
  34. 2026Reheat Nachos for Dinner? Evaluating AI Support for Cross-Cultural Communication of Neologismscited by 1
  35. 2026Priming, Path-dependence, and Plasticity: Understanding the molding of user-LLM interaction and its implications from (many) chat logs in the wildcited by 1
  36. 2026Do LLMs Hold Their Values? MANTA: A Multi-Turn Adversarial Benchmark for Animal Welfare Reasoningcited by 1
  37. 2026Going PLACES: Participatory Localized Red Teaming for Text-to-Image Safety in the Global Southcited by 1
  38. 2026Learning from 53.6K Real-World Developer Edits of AI-Generated Codecited by 1

50 most cited of 72 All 72 (JSON)