The goal of this project is to inject binary diversity into existing programs to lessen the ability of malware to infect multiple systems, much in the manner DNA diversity protects population from biological viruses. The research centers on finding ways to inject cryptographically secure randomization techniques into Intel instruction sets, at the function, process, or entire operating system level, with the resulting randomized binary being run on an emulation layer. The translation is designed to be automatic and transparent to end users. The end result will be each user can have some or all of their applications securely randomized in unique manners, making it vastly harder for many classes of malware to spread through a network on previously homogeneous computer systems.