How to Fund an Unfunded Artificial Intelligence Mandate

The federal government is often accused of not being able to keep policy aligned as new technologies emerge. Over the past two years, the administration has issued a lot of guidance around robotic process automation, or RPA, and the broader technology field of artificial intelligence. Whether sparked by China’s announcement of its intention to be the world’s AI superpower or just good strategic leadership, the emphasis is welcomed. While the latest executive order from the White House was immediately panned as an “unfunded mandate,” critics might consider the substantial leadership of several mid-level agencies before calling for funding in these fiscally challenging times.

Executive branch leadership over the past few years has been encouraging agencies to begin their path to AI, and Congress and agencies have begun that process. Five key directives have been issued to encourage government agencies to adopt an automation-first posture and ensure the United States continues its dominance in the AI field:

1. Office of American Innovation: Tasked with the mission to “make recommendations to the President on policies and plans to improve Government operations and services.”
2. Office of Management and Budget Memorandum 18-23: Recognizing that federal employees devote tens of thousands of hours to necessary (low-value, repetitive, rules-based) work the memorandum directs agencies to look for opportunities to shift workers to more value-added work by “using shared service solutions or technologies, eliminating agency-specific guidance or policies that preclude using shared services, and introducing new technologies, such as robotics process automation (RPA), to reduce repetitive administrative tasks, and other process-reform initiatives.”
3. 21st Century Integrated Digital Experience  Act: Sets in motion the opportunity for the consolidation of shared services across the government directing coordination with “other agencies and seek to maintain as much standardization and commonality with other agencies as practicable in implementing the requirements of this Act, to best enable future transitions to centralized shared services.”
4. The American Artificial Intelligence Initiative: This executive order, is designed to spur the development and regulation of AI by encouraging innovative government agencies to incorporate RPA and cognitive intelligence into their agencies.
5. Defense Department Artificial Intelligence Strategy: A strategy for not only laying out how AI will be ethically used as a battlefield multiplier but also how it will be used to improve the efficiency and effectiveness of Pentagon operations.

Michael Hermus, former chief technology officer at the Homeland Security Department, almost immediately called for agencies to “put existing money where their mouth is and actually make some changes” when speaking on a recent “Government Matters.” Forbes contributor Jessica Baron worries, “While an obvious concern is funding for these innovations, no announcements have been made about the specific financial resources that will become available to the new program.”

Critics could argue the United States is lagging behind 18 other countries that already have an AI strategy, but a strategy is not funding. Only nine countries are known to have funded AI strategies with China believed to have an estimated $150 billion allocated to become a global AI superpower by 2030. The gap between China and the next few countries in spending sees South Korea allocating $2 billion and Australia $25 million.

Certainly, China has thrown down the “robot gauntlet” but ingenuity and grit by a myriad of mid-level RPA practitioners across U.S. defense and civilian agencies offer immediate opportunities to seed AI research and development. More than 23 agencies began using RPA in 2018 and initial reports suggest savings are positive. For example, the Defense Information Systems Agency has piloted four RPA applications, including one that pulled documents for an audit faster than a human employee. The Defense Logistics Agency Information Operations automated parts of its onboarding process for an annual savings of $2 million and the General Services Administration uses six robots that shoulder work that took humans a combined 12,000 hours per year to complete.

Agency leaders can push to find those million-dollar savings automations, and then redirect staff and funds to other projects. The opportunity is there: Deloitte research suggests as much as $41 billion could be saved through the full implementation of artificial intelligence.

With significant executive branch support, agencies have the top cover to step on to the path of artificial intelligence by making investments today that are down payments for their future. Innovative and determined mid-level government employees have shown the path forward. It is time for leadership to challenge their staff to find work that can be automated and clear the way for a workforce that includes human employees and digital robots performing side by side.

Jim Walker is the former RPA lead at NASA Shared Services Center in Stennis, Mississippi. He is currently director of public sector marketing for UiPath.