How AI can affect productivity in public service delivery

What’s the issue? 

Artificial intelligence, and particularly generative AI technology, has the potential to enhance governance and public service delivery, significantly increasing the value of city-level programs. One important question involves how AI can support human public service discretion and while also enhancing and not sacrificing accountability.

What does the research say about how AI can improve public service delivery?

While more research is needed on how city governments can use AI-based tools in general, HKS Professor Stephen Goldsmith and Data-Smart City Solutions researcher Juncheng “Tony” Yang, find that a “balanced human-AI partnership” can create more efficient, responsive, and accountable public administration systems. The mere introduction of AI technology, the researchers say, calls for rethinking the traditional frameworks that guide public service administration.

Their paper addresses how the use of AI in city governments may help balance the traditional trade-offs these governments face between discretion and accountability. The authors emphasize that different types of public accountability—political, professional, and participatory—may be reshaped by AI, requiring tailored oversight mechanisms to address these shifts. In addition, Goldsmith and Yang explore the challenges of ensuring that AI systems are used ethically and offer some guiding principles for future AI integration. These guiding principles highlight the need for transparent data practices, accountable human-AI collaboration, and city-level governments’ role in mediating public perceptions of AI.

The authors discuss how AI tools can change the nature and quality of supervision which when combined with transparency will be keys to upholding public trust. In addition, they suggest that AI-trained government employees will be able to much more frequently use data to guide their service delivery which also will improve responsiveness and trust.