Disruption Opportunity

Human-AI Communication for Deontic Reasoning Devops (CODORD) intends to create new, automated techniques for humans to author knowledge about deontics (obligations, permissions, and prohibitions) into an expressively flexible logical language by using natural language (i.e., English). 

CODORD has the potential to enable automated deontic reasoning with high assurance (i.e., verifiability/explainability and correctness) to assess compliance with orders, regulations, laws, operational policies, and ethics. 

Automated deontic reasoning could be critical for enabling autonomous systems to:

  • Accurately apply rules of engagement in dynamic contexts
  • Interpreting and maintaining the integrity of commander's intent across the chain of command and in operations planning
  • Analyzing intelligence
  • Negotiating among subtle but crucial differences in legal and command authorities in allied relationships during joint operations.

Resources
Office
Defense Sciences Office
Program manager
Benjamin Grosof

Read bio

Opportunity

DARPA-PA-24-04-01
Solicitation

Accordion item
Performer papers

University of Southern California (USC)

  • Nananukul, N., Zhang, Y., Lee, R., Boxer, E., May, J., Gogate, V. G., Pujara, J., & Kejriwal, M. (2026). LOGicalThought: Logic-Based Ontological Grounding of LLMs for High-Assurance Reasoning. IJCAI 2026 Workshop on Logical and Symbolic Reasoning of Large Language Models. https://openreview.net/forum?id=60Kk4ms3Sz 

RTX BBN Technologies

  • Dou, G., Jurayj, W., Holzenberger, N., & Van Durme, B. (2026). DAR: Deontic reasoning with agentic harnesses. In preprint. https://arxiv.org/abs/2606.05009
  • Dou, G., Brena, L., Deo, A., Jurayj, W., Zhang, J., Holzenberger, N., & Van Durme, B. (2026). DeonticBench: A benchmark for reasoning over rules. In preprint. https://arxiv.org/abs/2604.04443 
  • Jurayj, W., Holzenberger, N., & Van Durme, B. (2026). Language Models and Logic Programs for Trustworthy Tax Reasoning. Proceedings of the AAAI Conference on Artificial Intelligence, 40(45), 38688–38698 https://doi.org/10.1609/aaai.v40i45.41212 
  • Jurayj, W., Cheng, J., & Van Durme, B. (2025). Is that your final answer? Test-time scaling improves selective question answering. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), 636-644. 2025. https://aclanthology.org/2025.acl-short.50.pdf

SRI International 

  • Cadigan, J., Wessel, M., & Morgenstern, L. (2026). Self-repair and explanation in automated large-scale translation of regulatory text to logic programming given partial or incorrect knowledge of context [Unpublished manuscript]. SRI International.

 

Contact