Skip to content

Argos: Securing the Information Access for AI Agents on Enterprise Relational Data

License

Notifications You must be signed in to change notification settings

tetherless-world/argos

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

30 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

ARGOS

Securing the Information Access for AI Agents on Enterprise Relational Data

Abstract

Traditional access control systems assume a deterministic, finite set of predefined data operations (actions), tightly coupling access rights to these actions. Retrieval-Augmented Generation (RAG) systems, however, connect large language models (LLMs) to enterprise databases, enabling dynamically generated, context-specific actions for which defining individual policies is both impractical and conceptually flawed. We introduce the ARGOS (Action Realignment using Graph-based Ontological Semantics), an ontology design pattern, which semantically represents actions, aligns them with structured data schemas, and governs access based on user-granted information operation boundaries. ARGOS employs a formal semantic policy framework with an extensible rule set to detect violations at the granularity of individual data operations and provide fine-grained, actionable feedback, thus moving beyond the conventional binary grant/reject paradigm toward context-aware, explainable enforcement.


Publication : ISWC WOP 2025 > Securing the Information Access for AI Agents on Enterprise Relational Data

Ontology

ontology

About

Argos: Securing the Information Access for AI Agents on Enterprise Relational Data

Resources

License

Code of conduct

Contributing

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 2

  •  
  •