Publications

The ethical, legal, and social issues (ELSI) surrounding Artificial Intelligence (AI) can have as great of an impact on the technologies’ success as technical issues such as safety, reliability, and security. Addressing these risks can counter potential program failures, legal and ethical battles, constraints to scientific research, and product vulnerabilities. This paper presents a surety engineering framework and process that can be applied to AI to identify and address technical, ethical, legal and societal risks. Extending sound engineering practices to incorporate a method to “engineer” ELSI can offer the scientific rigor required to significantly reduce the risk of AI vulnerabilities. Modeling the specification, design, evaluation and quality/risk indicators for AI provides a foundation for a risk-informed decision process that can benefit researchers and stakeholders alike as they use it to critically examine both substantial and intangible risks.