Augmenting SOAR with Deception Technologies for Enhanced Security and Application Response
Keywords:
Privacy by Design, Artificial Intelligence, Machine Learning, Differential Privacy, Cloud Computing, Data Security, User Privacy, AI/ML Systems, Secure Systems, Privacy-Preserving Techniques, Real-Time Data, Simulation, Data Protection, Privacy Measures, AI Ethics, Privacy Challenges, Secure AI Models, Privacy Solutions, Data Sensitivity, Computational EfficiencyAbstract
Incorporation of PBDe in AI/ML Systems is essential for the creation of secure and private environments, especially as data sensitivity continues to increase rapidly. This paper seeks to establish how to deploy secure AI/ML systems via the differential privacy approach on the cloud. In this way, the techniques' effectiveness is evaluated, taking into account real-time modes of work and using various datasets for the analysis. Our outcomes are depicted in enhanced graphs with the potency measures and consequence of privacy measures on a model's capability. It also discusses issues encountered while applying these techniques, including data utility and computational, among others, and how these issues can be effectively dealt with through some recommended and practical solutions. These findings emphasize the need to integrate effective privacy measures for protecting consumers' information processed by AI/ML technologies without compromising internal system performance and stability.
