Illuminating the Intersection of Institutionalized Artificial Intelligence Policies and the Use of AI in Business And Government Applications

Based on desk research and interviews with experts from government and the private sector, this report identified key obstacles that federal government agencies face in the development and implementation of AI systems. The analysis identified a range of different challenges public sector organizations face in AI adoption, which can be organized along three distinct but interrelated dimensions, namely Strategy (insufficient AI strategy, decentralized authority and expertise, unclear and indecisive leadership), Capabilities (human, technical and organizational resources), and Culture (disconnect between developer and user, impeded data sharing, risk aversion, fear for job replacement).  Based on this analysis and drawing on examples of good practice from both the United States and abroad, the report provided recommendations for organizational guidelines and policies that better enable public sector organizations to successfully develop and implement AI. The  recommendations provided include the necessity of strategy development first, building centralized and organic capabilities, enabling top-level leadership, centralized data structure, exchange of talent between public and private sectors, user centric development process, and roadmaps tailored to the unique needs of each organization to overcome cultural obstacles.