Abstract
The techniques presented herein implement an artificial intelligence (AI)-driven, risk-adaptive network automation system that determines how network configuration changes should be executed based on potential operational impact. The system interprets and classifies a natural language change request submitted by a human or generated by an autonomous agent. The system further evaluates real-time network context to calculate a composite risk score and a sphere of impact. Moreover, the system compares the score against user-defined risk tolerance thresholds to decide whether to execute the change autonomously, select a lower-risk alternative such as off-peak scheduling or staged deployment, or route the change to a human-in-the-loop change control process. By dynamically aligning execution strategy to risk, the system enables faster and safer automation while reducing the likelihood of disruptive network events.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Barton, Robert and Udupi, Yathi, "RISK-ADAPTIVE AUTONOMY FOR ARTIFICIAL INTELLIGENCE (AI)-DRIVEN NETWORK AUTOMATION SYSTEMS", Technical Disclosure Commons, (June 23, 2026)
https://www.tdcommons.org/dpubs_series/10554