Software

NASCIO 2021: States Reveal Accelerated AI Adoption in Survey

Agencies rapidly turned to artificial intelligence to handle large citizen service caseloads quickly, an upcoming report highlights.

Mickey McCarter

Mickey McCarter is the senior editor of StateTech Magazine.

“If you give me a lever and a place to stand, I can move the world,” said Archimedes in describing the power of mathematics.

Artificial intelligence will provide similar leverage for state and local governments looking to cut through red tape and process data, Virginia CIO Nelson Moe said Monday during the annual conference of the National Association of State Chief Information Officers in Seattle.

In Virginia, Moe plans to offer AI capabilities as a centralized service, bypassing a lack of skilled staff. His office will work with the commonwealth’s 65 agencies to adjust business processes as needed. The process will inform the Virginia Information Technologies Agency as to how well AI is adopted and its ROI.

Moe envisions popular AI use cases including predictive analytics, operations management, decision support, and the prevention of fraud, waste and abuse. “We want to take it up the value chain across the board,” Moe said in remarks during a panel titled “AI in State Government: Recent Impacts and Future Plans.”

Virginia will soon seek a contractor to help with AI rollout across government, he added.

“We are looking for entities that will help us with the back end of the AI and tool selection and with helping us understand the state of the market of what’s out there in AI, because it’s changing,” Moe said.

Upcoming AI Report Depicts Rapid Adoption by States

In the same panel, Bruce Tyler, a senior partner with IBM, presented takeaways from an annual AI report from the Center for Digital Government, IBM and NASCIO that is due for release soon.

In 2019 and 2020, the organizations collaborated on a survey of how state governments have adopted AI. The results from the 2021 survey reveal a stunning acceleration of AI adoption among states, according to Tyler’s preview.

This year, 60 percent of survey respondents said they have deployed AI solutions, compared with 13 percent in 2019. Further, 71 percent reported that AI improved service delivery, and 48 percent reported it enhanced interactions with constituents.

In 2019, “the numbers of states that were adopting AI were pretty minimal. We were in a pilot and exploratory phase,” Tyler said. “So, we quickly moved the needle from something that was exploratory to something that is widely adopted.”

The mood around AI changed during pandemic lockdowns as states deployed chatbots over the course of 2020, Tyler said. Many organizations were dealing with large caseloads and long lines, and AI helped resolve citizen demands. “AI had its moment in terms of pandemic response,” he added.

Sixty-five percent of survey respondents said AI is delivering results as expected. Only 10 percent indicated significant challenges in AI adoption.

While chatbots are a useful force multiplier, AI adoption quickly deepened across state enterprises, Tyler said. In the past 12 months, states deployed AI not only for digital assistants but also for robotic process automation, natural language processing, machine learning, computer vision, machine translation and more.

Survey Points to Expanded AI Use and Challenges to AI Adoption

“In a very quick time frame, AI moved from state governments trying to understand what it is to delivering results,” Tyler said. “I draw on my experience with cloud, and the cloud conversation has taken a decade to gain steam and adoption. So, for this to move so quickly is pretty impressive.”

State governments indicated further AI adoption to come, suggesting they would expand its use in all areas in the future. To facilitate such expansion, the survey found that states need a clear framework for AI use and governance, a defined vision and strategy for AI, and additional regulatory measures.

However, AI adoption also face challenges. Among the main bottlenecks in deploying AI are a lack of skilled staff, legacy modernization challenges, difficulty in identifying use cases and a lack of data or poor data quality, surveyed states said.

Check out more coverage from the NASCIO 2021 Annual Conference and follow us on Twitter at @StateTech, or the official conference Twitter account, @NASCIO, and join the conversation using the hashtag #NASCIO21.

Getty Images / Weedezign