A Novel Malicious Intent Detection Approach in Intent-Based Enterprise Networks

In an era dominated by digitalization in corporate strategies, the effectiveness of network infrastructures constitutes the foundation upon which enterprises build their businesses. Within this context, Intent-Based Networking (IBN) emerges as a networking paradigm that addresses the need for autonomous and intelligent systems, shifting network management from static to dynamic and automated processes. Despite its significant advancements in network automation, IBN introduces new complexities and challenges, particularly concerning the assessment of reliability or trustworthiness. Formulating intent without thorough investigation of users' intentions could potentially lead to deliberate network sabotage, denial of services, data breaches, and privilege escalations, thereby posing significant risks to enterprise networks. To bridge this gap, this study proposes a novel Malicious Intent Detection (MID) module within the IBN framework to construct a security knowledge base for accurately classifying enterprise users' intentions. Specifically, it detects malicious expressions directly during the intent processing stage, thereby preventing the formulation of disruptive network configurations or policies derived from malicious intents. The obtained results demonstrate the effectiveness of the proposed solution, with the ability to detect 94% of malicious intentions and achieve an accuracy of up to 93%.