Abstract:
Cybersecurity faces a constant increase in the development of techniques, malwares and cyber threats that exploit various vulnerabilities to achieve their objectives, presenting the potential to cause social and financial impacts for individuals and organizations. The complexity and diversity of cyber attacks make it difficult to create efficient responses, as for an attack to be effective it is necessary to exploit one or a few specific vulnerabilities, while to create a protected system it is necessary to implement defenses against diverse types of threats. Using the FLPN method to analyze the responses from ChatGPT, Bard and Perplexity and with the use of comparative graphs, this study’s main objective is to verify the accuracy and coverage of the responses of these artificial intelligences in comparison with the guidelines proposed by MITRE ATT&CK. In this work, ChatGPT achieved the highest average coverage score due to the number of items per response together with the average precision obtained. Therefore, it is possible to analyze the feasibility of cybersecurity analysts using generative artificial intelligence tools to respond to cyber incidents, contributing to the effectiveness of cyber defense strategies.
Referência:
COLLARES, Julio César; RICCARDI, Gianni; GAVA, Vagner Luiz. Estado da arte da inteligência artificial na resposta de incidentes cibernéticos. In: INTERNATIONAL CONFERENCE ON INFORMATION SYSTEMS AND TECHNOLOGY MANAGEMENT, 20, 2024, São Paulo. Proceedings… [on-line]
COLLARES, Julio César; RICCARDI, Gianni; GAVA, Vagner Luiz. Estado da arte da inteligência artificial na resposta de incidentes cibernéticos. In: INTERNATIONAL CONFERENCE ON INFORMATION SYSTEMS AND TECHNOLOGY MANAGEMENT, 20, 2024, São Paulo. Lecture and Abstract…
Acesso ao trabalho e resumo da Palestra em PDF:
https://www.tecsi.org/contecsi/index.php/contecsi/20thCONTECSI/paper/view/7214
