Gantral: Implementation of an Admissible AI Execution Control Plane
Year: 2026
Type: article
Abstract: Gantral is a reference implementation of an AI Execution Control Plane addressing operational inefficiency, operational fragmentation, and broken chain-of-custody in enterprise agentic AI systems. It binds authority to deterministic workflow state, separates policy from agent code, and emits cryptographically chained commitment artifacts enabling independent replay verification. This paper presents the formal specification and implementation architecture of the Gantral control plane.
Author: Phutane Babu, Abhishek Kumar
Language: English
Topic: Access Control and Trust
Subfield: Sociology and Political Science
Field: Social Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7130317714
FWCI:
Cites:
Cited by:
Related to: