TY - JOUR
T1 - Artificial Intelligence-Based Cybersecurity for the Metaverse
T2 - Research Challenges and Opportunities
AU - Awadallah, Abeer
AU - Eledlebi, Khouloud
AU - Zemerly, Mohamed Jamal
AU - Puthal, Deepak
AU - Damiani, Ernesto
AU - Taha, Kamal
AU - Kim, Tae Yeon
AU - Yoo, Paul D.
AU - Raymond Choo, Kim Kwang
AU - Yim, Man Sung
AU - Yeun, Chan Yeob
N1 - Publisher Copyright:
© 1998-2012 IEEE.
PY - 2025
Y1 - 2025
N2 - The metaverse, known as the next-generation 3D Internet, represents virtual environments that mirror the physical world. It is supported by innovative technologies such as digital twins and extended reality (XR), which elevate user experiences across various fields. However, the metaverse also introduces significant cybersecurity and privacy challenges that remain underexplored. Due to its complex multi-tech infrastructure, the metaverse requires sophisticated, automated, and intelligent cybersecurity measures to mitigate emerging threats effectively. Therefore, this paper is the first to explore Artificial Intelligence (AI)-driven cybersecurity techniques for the metaverse, examining academic and industrial perspectives. First, we provide an overview of the metaverse, presenting a detailed system model, diverse use cases, and insights into its current industrial status. We then present attack models and cybersecurity threats derived from the unique characteristics and technologies of the metaverse. Next, we review AI-driven cybersecurity solutions based on three critical aspects: User authentication, intrusion detection systems (IDS), and the security of digital assets, specifically for Blockchain and Non-fungible Tokens (NFTs). Finally, we highlight challenges and suggest future research opportunities to enhance metaverse security, privacy, and digital asset transactions.
AB - The metaverse, known as the next-generation 3D Internet, represents virtual environments that mirror the physical world. It is supported by innovative technologies such as digital twins and extended reality (XR), which elevate user experiences across various fields. However, the metaverse also introduces significant cybersecurity and privacy challenges that remain underexplored. Due to its complex multi-tech infrastructure, the metaverse requires sophisticated, automated, and intelligent cybersecurity measures to mitigate emerging threats effectively. Therefore, this paper is the first to explore Artificial Intelligence (AI)-driven cybersecurity techniques for the metaverse, examining academic and industrial perspectives. First, we provide an overview of the metaverse, presenting a detailed system model, diverse use cases, and insights into its current industrial status. We then present attack models and cybersecurity threats derived from the unique characteristics and technologies of the metaverse. Next, we review AI-driven cybersecurity solutions based on three critical aspects: User authentication, intrusion detection systems (IDS), and the security of digital assets, specifically for Blockchain and Non-fungible Tokens (NFTs). Finally, we highlight challenges and suggest future research opportunities to enhance metaverse security, privacy, and digital asset transactions.
KW - Artificial intelligence
KW - biometrics
KW - continuous authentication
KW - cybersecurity
KW - digital twins
KW - intrusion detection
KW - metaverse
KW - multimodality
KW - NFTs
UR - https://www.scopus.com/pages/publications/85201325073
U2 - 10.1109/COMST.2024.3442475
DO - 10.1109/COMST.2024.3442475
M3 - Article
AN - SCOPUS:85201325073
SN - 1553-877X
VL - 27
SP - 1008
EP - 1052
JO - IEEE Communications Surveys and Tutorials
JF - IEEE Communications Surveys and Tutorials
IS - 2
ER -