We need an AI strategy for government

Successful use of AI in government requires more than just a focus on compliance with controls.

Everyday there are multiple people writing that artificial intelligence has the potential to revolutionize government services, making them more efficient, responsive and citizen-centered. And on March 28, the Office of Management and Budget released the final guidance for agency use of AI (M-24-10). Having spent a long time at the highest levels of the legislative and executive branches of government and working with governments all over the world on how to effectively use technology, it is very clear that the new OMB memo has a major shortfall in guidance for using AI to improve government results.

AI has become the focus on strategy for both government and industry. Strategic efforts in government that are compliance-focused usually become reporting exercises that rarely improve agency results. While governance is a key component of any strategy, success in AI requires clear objectives and measurable outcome targets.

In fairness to OMB, it is hard to initiate strategy in the fourth year of an administration. As one of its first acts, the next administration should take a strategic approach to AI in government, emphasizing clear goals, measurable outcomes and a “know your customer” approach that has proven successful in the business world.

The public will best benefit from government AI initiatives tied to strategic needs and opportunities. For agencies, a widely understood vision and measurable targets are as important as the governance, risk and compliance requirements crafted by the Office of Management and Budget. Based on what we know about the agency AI initiative submission to the White House last year, about a thousand AI projects are underway in government. Those projects range from good old fashioned algorithmic AI (aka GOFAI) to robotic process automation (RPA) to neural networked generative AI tools and chatbots. We should not spend billions of dollars on AI projects that are compliant with OMB guidance and are merely interesting from a research perspective. Such a field of dreams approach has rarely succeeded in improving government. Rather, with so much effort on AI projects, we need a return on investment tied to the biggest strategic needs and chronic problems of government.

Anyone who has watched a business show or tracked technology stocks over the past nine months is aware that AI has become a necessity for companies. One approach that garners corporate attention is the use of AI in customer relationship management (CRM) for understanding and supporting customer needs. Take, for example, companies that provide myriad products and services, such as Amazon and Facebook Marketplace. They know you as one customer who has unique needs, and they customize offers for those needs. Even beyond that, interactions are simplified, and data rarely must be re-entered.

The government could improve efficiency, effectiveness and responsiveness from using this “know your customer” approach.

For example, AI tools can predict when participants are likely to need recertification for benefits or when they might benefit from other support services, and proactively reach out to them with relevant information and aid.

Similarly, as the Pandemic Analytics Center of Excellence has documented, this approach can protect beneficiaries and taxpayers against fraud, waste and abuse.

By analyzing patterns in participant feedback and complaints, AI algorithms can pinpoint areas where programs are falling short and suggest targeted interventions to improve outcomes. For example, if a program receives negative feedback or does not achieve outcome targets, AI can illuminate the root causes of the problem and propose solutions to enhance the program’s effectiveness. While this can lead to faster response times, prevent fraud, improve satisfaction, and ultimately, more efficient program operations, it still may only lead to incremental change.

I learned during my time on Capitol Hill and at OMB that substantial improvements require major changes in programs. So if companies can use AI to “know your customer” and provide an efficient integrated experience across multiple service and product providers, why should not the government apply the same approach to address major problems and governmentwide strategic needs?

Insights from Professor Henry Mintzberg, author of The Rise and Fall of Strategic Planning, provide a useful approach for strategic considerations of AI. Mintzberg wrote that strategy needs to be holistic, addressing the organization’s culture, structure and processes. He viewed strategy as a way of thinking and learning about challenges and opportunities.

So a planning process that overlays GOFAI, generative AI and RPA opportunities with existing government strategies, learning agendas and OMB’s controls could drive government agencies to consider AI initiatives through the lens of government performance breakthroughs.

A proven approach, such as the Quick Silver framework using a collaborative, cross-agency team lead by OMB or the President’s Management Council would go far in moving from the current AI compliance guidance to an outcome-driven strategy.

Such an approach would generate a governmentwide strategy for AI. Agency priority goals and cross agency goals would incorporate AI-infused concepts (e.g., know-your-customer) as a key component for significant improvements to customer service and program effectiveness. AI would be integrated into this strategy as a tool to support the achievement of strategic goals, and AI initiatives would be funded in alignment with the strategic goals.

The benefit here, as we found in the Bush administration’s e-Government effort, is that the collaborative effort spawns holistic inter- and intra-agency innovation as the team members return to their home agency. Applying a “Quick Silver” approach would create a shared vision and understanding of how AI fits within a comprehensive approach to enhance overall government performance and service delivery, mitigating the limitations and risks of a siloed agency-by-agency approach while following the desired controls in the current guidance.

Successful use of AI in government requires more than just a focus on compliance with controls. There must be a vision and approach for achieving strategic value. It is the only way to assure government performance gains as well as that the guidance does not get reduced to a reporting exercise. Applying AI using commercial “know your customer” concepts to federal programs holds the potential to streamline processes, clarify which programs are effective and which are not, and ultimately lead to more effective and efficient program operations. However, it is essential to have clearly said goals and shared vision across agencies to ensure that the benefits are realized without compromising individual rights and societal values.

Mark Forman is the first administrator for e-government and IT in OMB and now executive VP at Dynamic Integrated Services.

Copyright © 2024 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.

Related Stories