Putting AI ethics to work: are the tools fit for purpose for SMEs?

Bias, unfairness and lack of transparency and accountability in Artificial Intelligence (AI) systems have raised concerns about the ethical impact, and unintended consequences of new technologies for society across every sector where data-driven innovation is taking place. This thesis first reviews the landscape of proposed ethical frameworks with a focus on those which go beyond high-level statements of principles and offer practical tools. It then provides an assessment of these practical ethics tools through the lens of known best practices for impact assessment and audit of technology. It reviews other historical uses of risk assessments and audits to create a typology that allows us to compare current AI ethics tools to best practice found in previous methodologies from technology, environment, privacy, finance, and engineering. It analyses current AI ethics tools and their support for diverse stakeholders and components of the AI development and deployment lifecycle. Building on this analysis, a series of interviews were conducted with CEO’s/founders of smaller tech companies to understand how these tools might be used (or not) in the production of real products and services. This uncovers a narrower conception of ethical concerns and stakeholders in the sector than presented in AI ethics tools and principles. The sector also understands itself as already taking the necessary steps to address ethical issues without the need for specific ethical tools or governance. From this, gaps are identified in current AI ethics tools and their practical application that should be considered going forward