AI technology could reshape the U.S. government, but should it?

Artificial intelligence is already making its mark on the global economy, and private businesses are not the only ones putting it to good use. Government agencies of all types are adopting AI as well, and in some ways,  exceeding the private sector’s ability to leverage AI for massive data analysis and cutting-edge applications.

Whether this sounds like innocent adoption of new technology or a nefarious plot to control the citizenry depends on your political perspective, of course. However, there is no denying that the same technology that is currently powering the likes of giants like Facebook and Amazon to leverage user information in pursuit of profits, is also available to all aspects of government, including taxation, defense, intelligence agencies, and other key entities like agriculture and labor.

What can AI do for the government?

According to a recent report by Deltek, federal spending on AI in the U.S. increased by 50% between 2018 and 2020, reaching nearly $1 billion — making it the fastest rate of growth for any emerging technology. While AI was once the purview of science-facing agencies like NASA and the Department of Energy, this technology is now migrating across the governmental spectrum in the quest to improve performance, create operational efficiencies, and reduce waste, fraud, and abuse. In large part, this transformation is being driven by legislation and policy directives from the highest levels of government.

So, what exactly can AI do for the government? Plenty, according to Boston Consulting Group’s Nadim Abillama, Steven Mills, Greg Boison, and Miguel Carrasco. For one thing, it can foster better policymaking by giving decision-makers more accurate and timely information regarding demographics, behavioral trends, and a wide range of other metrics. At the same time, it can quickly provide the kind of analytical feedback needed to determine if a particular policy or program is performing as intended, and at the anticipated cost level.

AI can also reinvent the (often dismal) user experience with government agencies. Ranging from chatbot-driven self-help tools to more customized interactions based on personal histories, socioeconomic factors, eligibility requirements, and a host of other data sets. Anyone who has tried to navigate the labyrinth of rules and regulations at, say, the Internal Revenue Service might appreciate help from any intelligence, whether it is artificial, biological, or otherwise. In many cases, such as with the Veterans Administration or Medicare/Medicaid, the performance improvements could be life-saving even as the cost reductions reach into billions of dollars.

Local Help

But it’s not just the federal level that is wading into the AI waters. State and local governments are getting their feet wet as well. San Leandro, California, a city with a population of 90,000, recently installed a platform called CityDash that provides a unified, intelligent data visualization framework for a wide range of municipal services. The system provides tools like a mobile chatbot for sharing data on everything from crime incidents, to building permits, and community events. CityDash also features a cloud-based knowledge graph with machine learning capabilities for analyzing IoT datasets relating to traffic flow, utilities, and even the weather. There is also a public-facing dashboard that assists citizens with non-emergency services and general inquiries.

AI is already fueling the development of smart cities that are expected to produce all manner of societal benefits, says A.J. Abdallat, CEO of AI development firm Beyond Limits. With better traffic management, for example, we should see lower carbon emissions, fewer traffic jams, fewer accidents, and improvements to infrastructure development and repair. At the same time, cities should be able to manage utilities like water and electricity, as well as services like garbage collection and public safety, in a more finite manner, directing resources where they are needed most and reducing waste or duplicative efforts. And this won’t just improve local operations but also the often tricky coordination that takes place between local, state, and federal authorities as well as quasi-state and private entities.

Government is in a unique position when it comes to AI in that it is deploying it at the same time it seeks to regulate it. While transparency has been one of the hallmarks of recent efforts to place some control on how businesses utilize AI, certain segments of government, particularly on the federal level, might not be so open. That will only breed suspicion as to what agencies are doing with our data and how it is being studied and manipulated.

At a time when mistrust of both the government and AI is running high, the idea that somewhere someone in authority is using AI behind closed doors will only increase public unease, regardless of how much the technology is improving government services.