A Secure API-Driven Framework for AES Modes of Encryption Enhanced with Machine Learning

Cryptography is a field within Computer Science and Mathematics that deals with various techniques of enciphering and deciphering data. One of the standard classes of algorithms used today in cryptography is called Advanced Encryption Standard (AES). Although these algorithms provide sufficient strength, they are not unbreakable, either by the potential of quantum computing or adversaries trying to decipher data using various attacks. This study shows the potential of using machine learning (ML) by building an application Programming Interface (API) driven framework for encryption and decryption that utilizes ML to add abstracted layers of security over encrypted data. Techniques such as Encapsulation and Abstraction that are used on the ciphertext and key, can only be interpreted by the developed framework. The utilization of this framework will ultimately provide a cohesive endend system that can limit many vulnerabilities and prove how ML can be utilized as a tool against cyber-attacks.