mlsanityData to 7 Oct 2026

Structured access: an emerging paradigm for safe AI deployment

arXiv 2201.05159 ↗First posted 13 Jan 2022cs.AI cs.HC cs.SEcited by 40

Not on the map. Why?

Abstract (arXiv)

Structured access is an emerging paradigm for the safe deployment of artificial intelligence (AI). Instead of openly disseminating AI systems, developers facilitate controlled, arm's length interactions with their AI systems. The aim is to prevent dangerous AI capabilities from being widely accessible, whilst preserving access to AI capabilities that can be used safely. The developer must both restrict how the AI system can be used, and prevent the user from circumventing these restrictions through modification or reverse engineering of the AI system. Structured access is most effective when implemented through cloud-based AI services, rather than disseminating AI software that runs locally on users' hardware. Cloud-based interfaces provide the AI developer greater scope for controlling how the AI system is used, and for protecting against unauthorized modifications to the system's design. This chapter expands the discussion of "publication norms" in the AI community, which to date has focused on the question of how the informational content of AI research projects should be disseminated (e.g., code and models). Although this is an important question, there are limits to what can be achieved through the control of information flows. Structured access views AI software not only as information that can be shared but also as a tool with which users can have arm's length interactions. There are early examples of structured access being practiced by AI developers, but there is much room for further development, both in the functionality of cloud-based interfaces and in the wider institutional framework.

Key facts. Cited by 40 papers in the map (data through 7 Oct 2026). Builds on 9 of its 27 references that are in the map. First posted on arXiv 13 Jan 2022.

Builds on

9 of 27 references.

References · 9

  1. 2020Language Models are Few-Shot Learnerscited by 32,199
  2. 2019DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generationcited by 842
  3. 2016Stealing Machine Learning Models via Prediction APIscited by 602
  4. 2019Release Strategies and the Social Impacts of Language Modelscited by 505
  5. 2020AI Research Considerations for Human Existential Safety (ARCHES)cited by 33
  6. 2021AdvParams: An Active DNN Intellectual Property Protection Technique via Adversarial Perturbation Based Parameter Encryptioncited by 11
  7. 2020Deep-Lock: Secure Authorization for Deep Neural Networkscited by 10
  8. 2018DeepObfuscation: Securing the Structure of Convolutional Neural Networks via Knowledge Distillationcited by 6
  9. 2019MimosaNet: An Unrobust Neural Network Preventing Model Stealingcited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 1; 2023: 15; 2024: 12; 2025: 5; 2026: 7.

40 papers.

Sort:

By area:

  1. 2024The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearningcited by 488
  2. 2023Model evaluation for extreme riskscited by 167
  3. 2024Black-Box Access is Insufficient for Rigorous AI Auditscited by 123
  4. 2024Large Language Model Unlearning via Embedding-Corrupted Promptscited by 112
  5. 2023Frontier AI Regulation: Managing Emerging Risks to Public Safetycited by 95
  6. 2024On the Societal Impact of Open Foundation Modelscited by 64
  7. 2023The Gradient of Generative AI Release: Methods and Considerationscited by 63
  8. 2023Spear Phishing With Large Language Modelscited by 49
  9. 2024The Oscars of AI Theater: A Survey on Role-Playing with Language Modelscited by 47
  10. 2024NNsight and NDIF: Democratizing Access to Open-Weight Foundation Model Internalscited by 43
  11. 2023Open-Sourcing Highly Capable Foundation Models: An evaluation of risks, benefits, and alternative methods for pursuing open-source objectivescited by 31
  12. 2023Democratising AI: Multiple Meanings, Goals, and Methodscited by 29
28 more
  1. 2024The Compute Divide in Machine Learning: A Threat to Academic Contribution and Scrutiny?cited by 20
  2. 2024Generative AI Misuse: A Taxonomy of Tactics and Insights from Real-World Datacited by 20
  3. 2024Holistic Safety and Responsibility Evaluations of Advanced AI Modelscited by 18
  4. 2023A Cost Analysis of Generative Language Models and Influence Operationscited by 17
  5. 2023Protecting Society from AI Misuse: When are Restrictions on Capabilities Warranted?cited by 15
  6. 2023Towards Publicly Accountable Frontier LLMs: Building an External Scrutiny Ecosystem under the ASPIRE Frameworkcited by 11
  7. 2024What is it for a Machine Learning Model to Have a Capability?cited by 10
  8. 2024Societal Adaptation to Advanced AIcited by 10
  9. 2025The 2025 Foundation Model Transparency Indexcited by 9
  10. 2023International Governance of Civilian AI: A Jurisdictional Certification Approachcited by 8
  11. 2022Actionable Guidance for High-Consequence AI Risk Management: Towards Standards Addressing AI Catastrophic Riskscited by 7
  12. 2023Hazards from Increasingly Accessible Fine-Tuning of Downloadable Foundation Modelscited by 7
  13. 2023Exploring the Relevance of Data Privacy-Enhancing Technologies for AI Governance Use Casescited by 5
  14. 2023An International Consortium for Evaluations of Societal-Scale Risks from Advanced AIcited by 5
  15. 2023Towards Responsible Governance of Biological Design Toolscited by 5
  16. 2025An Example Safety Case for Safeguards Against Misusecited by 4
  17. 2025Technical Requirements for Halting Dangerous AI Activitiescited by 2
  18. 2026Internal Deployment Gaps in AI Regulationcited by 2
  19. 2026Position: Behavioural Assurance Cannot Verify the Safety Claims Governance Now Demandscited by 2
  20. 2023Contextual Confidence and Generative AIcited by 1
  21. 2025Beyond Release: Access Considerations for Generative AI Systemscited by 1
  22. 2026A Content-Based Framework for Cybersecurity Refusal Decisions in Large Language Modelscited by 1
  23. 2026The Ethics of Autonomous AI Agents for Offensive Securitycited by 1
  24. 2024Considerations Influencing Offense-Defense Dynamics From Artificial Intelligencecited by 0
  25. 2025Recognition Without Mitigation: Ethical Frameworks in Autonomous Offensive-LLM Agent Researchcited by 0
  26. 2026A Privacy by Design Framework for Large Language Model-Based Applications for Childrencited by 0
  27. 2026Incentive Aware AI Regulations: A Credal Characterisationcited by 0
  28. 2026Agentic BAIM-LLM Evaluation (ABLE): Benchmarking LLM Use of Protein Design Toolscited by 0