8Edge AI Security and Privacy: Threats, Solutions, and Best Practices
Ishika Mohan1, Advika Wankhede1, Ishita Manral1, Rohit Kadam1, Preeti Agarwal1* and Anchit Bijalwan2
1School of Technology Mangement & Engineering, Narsee Monjee Institute of Management Studies (Deemed to-be University), Navi Mumbai, Maharashtra, India
2School of Computing and Innovative Technologies, British University Vietnam, Hanoi, Vietnam
Abstract
Edge AI, through the requirement of real-time processing of data closer to the source, brings unique challenges in security and privacy due to its decentralized architecture and deployment in uncontrolled environments. This chapter explores various threats and vulnerabilities in edge devices, including physical tampering, adversarial attacks, model poisoning, and insecure communication protocols. It underlines the risks related to resource constraints, outdated software, and regulatory compliance necessities like GDPR and HIPAA. Advanced threats, including malware, DDoS, and side-channel attacks, are analyzed with a focus on implications for data integrity and system reliability. The emerging solutions discussed include federated learning, homomorphic encryption, and blockchain as mechanisms to mitigate these risks. This chapter looks at best practices, including the use of strong encryption, AI-driven threat detection, and zero-trust architectures, to improve the security and privacy of edge systems. By using innovative and interdisciplinary approaches to ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access