Abstract
Proposed herein is a Network Agent Policy Language (NAPL) that provides a declarative policy language and a nine-component enforcement architecture for governing the actions of autonomous artificial intelligence (AI) agents operating on network infrastructure. As AI agents gain the ability to invoke network management Application Programming Interfaces (APIs), configuration tools, and telemetry interfaces in real time, existing policy frameworks designed for human operators or stateless access control cannot express or enforce the governance rules that network operations require. NAPL addresses this issue by introducing network-domain semantics directly into policy authoring, enabling operators to write governance rules using primitives such as bgp_neighbors, ospf_areas, vlan_ranges, and interface_classes rather than low-level configuration paths or generic resource strings.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Jeuk, Sebastian; Simon, Katrin; and Haasch, Jan, "NETWORK AGENT POLICY LANGUAGE (NAPL)", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/10273