What government CIOs need for AI to succeed

(Getty Images)

Share

 
Written by

Kirke Everson is a principal in KPMG’s Federal Advisory practice, focusing on technology enablement, intelligent automation, program management, process improvement, cyber security, risk management, and financial management. He currently serves as the government lead for Intelligent automation for KPMG in the U.S.

artificial intelligence
Kirke Everson, Principal, Federal Advisory, KPMG

Federal and state government leaders are witnessing the expansion of artificial intelligence all around them. From back-office automation, that can help reduce backlogged work, to cognitive platforms, that can identify and respond to natural language requests to better serve the public, AI and automation have become a driving force in addressing mission and business objectives.

That’s clearly evident in speaking directly with federal and state government CIOs in multiple roundtable discussions over the past several weeks. Based on the use cases they described, it’s clear that agencies are making significant headway in putting AI to work.

At the same time, there are a variety of issues where government CIOs also need broader support. The issues they and their executive teams face, in many ways, are not that different from previous technology breakthroughs that tended to upend familiar work processes. The technology component — like the disruption of mobile and cloud technology — is only part of a larger equation involving processes, policy, culture, governance and ethics. That said, government is already seeing the value of AI, especially in light of unprecedented citizen demands for agency services during the pandemic.

CIOs in these roundtable discussions expressed a collective optimism and determination for how government can and must put AI to work. There was a broad consensus that they cannot delay integrating AI into their operations. “It’s not a question of when, but how we allow AI to come into our processes,” said one leader.

What’s critical now for agencies is to address these larger policy issues. At the same time, they also need to assess and select the right technologies for more advanced use cases; establish the means to scale AI solutions; and improve the quality of, and access to, clean and digestible data.

It’s getting to the “how” that CIOs are now wrestling with. Here’s a partial list of what many of them say their organizations still need from their leadership to ensure AI will live up to its promise:

Clearer use cases — We have already seen how AI is helping government agencies to analyze data, automate responses through robotic process automation and augment employee workloads. But agencies have all kinds of opportunities to build upon those successes to develop more advanced use cases, where more complex AI components can be applied to drive innovation and strategic decision making for better mission outcomes. Fortunately for government, AI’s application in commercial sectors — for detecting fraud, for example — are demonstrating tangible results and offer a helpful roadmap for what’s possible.

Greater data preparation The ability to readily identify useful patterns in government data depends on having clean and reliable data to work with. Given the volume and velocity of data agencies generate, CIOs need modernized IT infrastructure, more robust data management applications and the resident skills to capitalize on them. Here is where automation and AI solutions can be part of the solution, to help access, consolidate, normalize and cleanse data.

Commitment to trustworthiness This was one of the top concerns among CIOs in these discussions. It is vital that senior leadership establish and implement an ethical and responsible approach to AI, that adequate controls be put in place, and that agencies execute against an ethical AI framework. Without those safeguards, agencies run the risk of relying on untrustworthy or biased data or undermining the productivity gains AI promises.

Greater governance — Decisions need to be made on how agencies will manage the application of AI, which use cases to pursue and how to implement, scale, monitor and evaluate the impact of AI. CIOs can’t do that alone. Governance will also play a key role in building in proper checks and balances and guidance on how AI is ultimately put to work — and ensuring that AI initiatives conform with wider accepted practices.

Broader training Artificial intelligence is more than the technology that creates it. It is a powerful set of outcomes that requires care, calibration and control — beginning with the data that goes into it and for the decisions that come out of it.  Not unlike the fields of medicine or building design, AI will require specialized and continuous training at many levels across every organization endeavoring to integrate AI into its operations.

There’s no question, AI promises to be an incredible force-multiplier in helping government better serve the public. CIOs have a unique perspective and a critical role to play in all of these areas.  But AI’s promise also hinges on agency leaders focusing on the mission or business, developing clear use cases, and then applying the technology, not the other way around.

Learn more about how KPMG can help your organization harness AI more effectively.

-In this Story-

artificial intelligence (AI), Digital Transformation, executive perspective, IT Modernization, Kirke Everson, KPMG, KPMG 2021, Sponsored Content
TwitterFacebookLinkedInRedditGmail