• Oladoyin Akinsuli

The increased rate of AI development and its applications across different fields have led to an important call for security frameworks to safeguard AI data and uphold the AI systems’ trustworthiness. Lacking types of security models that would serve well the increasing threats of AI technologies, new concepts of solutions have to be proposed, like the Zero Trust Security Architecture one (ZTSA). This present paper looks at how Zero Trust security model can be used in helping protect the Artificial Intelligence systems; specifically, the model’s tenet of constantly checking all entities and the assumption those threats could emerge from both internal and external sources. Zero Trust Security Architecture is grounded on the principle of “do not trust” or “verify,” and it eradicates preconceived trust in the organization’s network systems. In the context of AI, it means proper access control, real-time surveillance, and rigorous security measures to safeguard data transfer line and the procedures within the algorithms. From the above-detailed ZTSA, organizations can protect AI systems from various threats such as data leakage, adversarial attacks, and unauthorized access of valuable resources. Consequently, this research explores the different elements of ZTSA including micro-segmentation, IAM, MFA, and sophisticated encryption techniques with reference to the AI. Since micro-segmentation provides a way of establishing logical subnets of a network, it limits the area that an unauthorized user can compromise and move around within the network. IAM makes sure that all the users who want to gain access to AI resources are properly identified and approved while MFA doubles down on the above by making it very hard for an attacker to get past a credential’s theft. Encryption is another area that should be very sophisticated because of data that has to be both transmitted and stored securely. In addition, the paper also depicts the difficulties and strategies for the realization of ZTSA for AI systems that includes some of the issues of security management that are important to embrace including the need to monitor implementation from time to time while incorporating security measures to adapt to new kinds of threats. This calls for proactive security management, complemented by the use of artificial intelligence-based analytical tools in threat analysis.

Zero Trust Architecture (ZTA), Artificial Intelligence (AI), Cyber security, Data Security, Network Security, AI Security, Machine Learning Security, Deep Learning Security, Model Security, Data Privacy, Access Control, Identity and Access Management (IAM), Micro segmentation

"Using Zero Trust Security Architecture Models to Secure Artificial Intelligence Systems", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 4, page no.q349-q373, April-2024, Available :http://www.jetir.org/papers/JETIR2404H46.pdf

"Using Zero Trust Security Architecture Models to Secure Artificial Intelligence Systems", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 4, page no. ppq349-q373, April-2024, Available at : http://www.jetir.org/papers/JETIR2404H46.pdf