Abstract
The Fifth Generation Standard (5G) allows for private actors to deploy self-owned telecommunication infrastructures. These private 5G deployments may be deployed in a broad range of contexts with differing requirements, increasing the need for resources to sufficiently optimize and secure specialized deployments. We introduce a framework to enable Generative Artificial Intelligence (GenAI) to be used to detect vulnerabilities in 5G Radio Access Network (RAN) deployments, in order to automate security testing activities. We implement a proof-of-concept deployment, detecting and verifying a vulnerability in a 5G-RAN deployment.