Threat Identification using STRIDE and GPT based chatbots
Threat Identification using STRIDE and GPT based chatbots

Summary/Abstract: This study demonstrates the effectiveness of using Large Language Models, such as ChatGPT, for the purpose of identifying potential threats and proposing effective defensive measures for protecting critical network services. Modern systems consist of multiple workloads of diverse types, which makes it challenging to manage the process of identifying threats manually, Artificial Intelligence can be utilized to assist with the chore of identifying threats. Using a comparison methodology a comparative analysis of three distinct GPT based tools (ChatGPT 4.0, STRIDE-GPT and custom gpt-4-1106-pre- view based model) that were assigned the responsibility of identifying threats based on the Nmap reports was conducted. The results demonstrate the feasibility of employing ChatGPT to aid in threat detection. Furthermore, they indicate that the risks recognized by AI are comparable to those identified by human experts, while also being delivered significantly faster when utilizing AI.

• Details
• Contents