SAGE: A systematic approach to data-driven AI governance

Artificial intelligence (AI) technology is advancing at an unprecedented rate, increasing in complexity while driving significant innovation across sectors. In response to the rapid development of AI, myriad compliance frameworks are being established to ensure the responsible development, deployment, and use of this emerging technology.

These frameworks include regulatory schemes, newly enacted laws, and a growing body of industry and technical standards, each with the goal of furthering safety, trustworthiness, and fairness. Providers, developers, and deployers of AI systems are acutely aware of the need to implement robust controls around their interactions with AI – not only to comply with the law, but also to maintain public trust, mitigate litigation risk, and ensure they are using AI responsibly.

In this paper, we will begin with an overview of the compliance landscape and then describe our method of building a comprehensive methodology – the Standardized Assessment and Governance Enhancement (SAGE) framework – to address these problems. Lastly, we will describe how we leveraged this novel methodology to provide companies with actionable guidance to understand and fulfill their obligations under the developing compliance landscape.