Conversational Artificial Intelligence

19Security Threats and Security Testing for Chatbots

Domenic T. Sanchez^1* and Rodel S. Sartagoda²

¹College of Education, Cebu Technological University-NEC, Cebu City, Philippines

²Department of Education - Cebu City Division, Guadalupe, Cebu City, Philippines

Abstract

Chatbots are vulnerable to vulnerabilities and attacks that are often utilized, such as cross-site scripting and SQL injections. As a result of these variables, it is quite probable that chatbots will also be the target of an attack. As a result of this, it is vital, while analyzing chatbots, to examine for problems about their level of security. The process of scanning and assessing vulnerabilities is a thorough analysis of computer networks and the components that make up such networks. By doing this study, we want to get a better understanding of the safety precautions that are now in place as well as the degree to which they are effective. The fundamental objective of penetration testing is to locate vulnerabilities in a protected setting so that they may be patched up before they are exploited by malicious actors (hackers, for example). In addition, doing penetration testing may be of use in determining how well the system can withstand genuine assaults. This article presents an investigation of vulnerability assessment and penetration testing for chatbots.

Keywords: Chatbot security, privacy, vulnerability assessment, penetration testing, vulnerabilities, attacks, threats

19.1 Introduction

Like any other ...