Resolving ambiguities in regulations: Towards achieving the kohlbergian stage of principled morality

Abstract

According to Kohlberg, the final stage of morality is characterized by viewing laws as a means to an end by upholding values such as human dignity and fairness as guiding principles for complying with the essence of the law. Given that purpose of compliance is indeed wellbeing of citizens, software systems should, by design, incorporate these values so that laws are followed in spirit. How can we build software systems that incorporate these values? We present our work on disambiguating Health Insurance Portability and Accountability Act (HIPAA) so as to reduce the potential incidents of breach, thereby upholding of the aforesaid guiding principles of morality. We have employed deep learning based approaches to emulate the human process of disambiguation by integrating information from multiple sources, summarizing it, and augmenting the regulatory text with the additional information. This augmented regulatory text can be used by policy makers and software engineers to achieve compliance in spirit.