Ep. 1: Where Government Should Begin with AI

No items found.

Governments are under growing pressure to adopt AI, often before the problem it is meant to solve has been defined. In this opening episode of Human Terms, Victoria Aarons interviews CIPI’s GovTech Expert in Residence, Mahesh Nattanmai, a former Chief Information Officer for New York State, about where the public sector should begin. Their discussion examines what efficiency means in government, why efficiency alone is an insufficient justification for adoption, and how the stakes change once AI systems make or shape decisions about people’s lives.

Drawing on Mahesh’s experience of leading technology inside government, the conversation considers how agencies can distinguish a real problem from a fashionable solution, and who bears the cost when automated systems fail, particularly people who already face barriers to work and public services.

The episode accompanies our policy brief, Where Government Should Begin with AI, which argues that the benefits of AI are real but conditional. Each deployment should begin with a defined problem, demonstrate evidence of value and carry safeguards proportionate to the stakes. The brief concludes with a practical test and a set of questions to ask before AI is put to use.