On Thursday, October 30th, the IBM Center for the Business of Government published a new report, “AI In Government: Balancing Innovation, Efficiency, and Risk.” 

It was researched and written by the proprietors of this website, Katherine Barrett and Richard Greene, and relied on interviews with many experts in the field, including Doug Robinson, executive director of the National Association of State CIOs (NASCIO),  Alan Shark, co-chair of the National Academy of Public Administration’s Standing Panel on Technology Leadership and the CIOS of a number of states including California, Connecticut, Massachusetts, Ohio. Texas and Utah.

Much of the current use of AI is for internal facing Chatbots, which can help public sector employees to do their jobs more efficiently, but as the report points out other ways to utilize AI are now emerging. Use Cases are included in the report about the practical use of AI in seven states, including New Jersey, Pennsylvania, Virginia Washington, and Wisconsin. 

Washington, for example, according to the report, is using AI to fight wildfires. As James Galvin, the lead for Washington’s AI and Emerging Tech Program, told us, “Today, wildfire detection helps fill that gap with 30 high powered cameras on tall towers looking across the state, helping human operators in identifying potential wildfires. They’re using AI in these tools to more accurately identify what is genuinely wildfire smoke versus dust from a dust storm or low hanging clouds. They’re able to continually improve the algorithms. So, using AI, they’re able to really improve the quick identification of wildfires and have less false positives.”

The report includes key elements that contribute to the successful use of AI, while avoiding pitfalls. For example: 

• “Be certain that the existing processes being used in government—like gathering accurate data—are functioning well, before utilizing them to feed AI systems.”

• “Put the time and money necessary into adequate training for government employees at all levels—and the public as well.” 

• “Whatever governance structures are put into place, make certain that they are being followed, and that the necessary funding is available to do so.” 

• “Ensure genuine transparency before embarking on an AI project to preserve accountability and the principles embodied in the Freedom of Information Act.”

• “Evaluate the benefits of AI. This goes beyond simply measuring speed and efficiency, and extends into measurements of fairness, equity, and the trust of the populace that AI is serving them well.” 

• “Be clear about what AI cannot do (at least yet). Extending its use in ways that don’t add value is a risk at a time when AI appears to be a panacea to a whole host of government problems.”

Beyond these warnings, many clear observations about the use of AI in the states are covered by the IBM Center. Among them are: 

• “While there’s a level of fear about humans being replaced by AI, the most commonly repeated statement in interviews was the vital importance of keeping ‘humans in the loop.’”

• “Though GenAI has been getting tons of attention, it’s clear that many states are still in the early stages of implementation. Pilots abound, but taking the programs to scale is less common than might have been anticipated.”

• “It’s still unclear as to what impact AI will have on the number of employees on state payrolls—though strong arguments can be made that at least some specific jobs will disappear.”

• “Though most states are aware that they need governance over their agencies’ use of AI, the approaches vary from very constrictive controls to greater flexibility.”

• “The most common use of GenAI currently is in a variety of chatbots, which can help to answer questions at lightning speed. However, so far much of the sophisticated use of these tools is for internal purposes and not outward facing for public use.”

#StateandLocalAI #StateandLocalArtificialIntelligence #StateManagementArtificialIntelligence #StateArtificialIntelligenceUseCases #StateAIUseCases #StateGovernmentGenerativeAIPolicyAndManagement #StateAIEmployeeImpact #StateArtificialIntelligenceTraining #StateArtificialIntelligenceGovernance #StateandLocalGovernmentData #StateDataGovernance #StateDataUse #NASCIO #DougRobinson #AlanShark #BarrettandGreeneStateArtificialIntellligenceReport #IBMCenterBusinessOfGovernment #StateAIControlVSFlexibility #StateArtificialIntelligenceImpact #AIandHumansInTheLoop #StateEvaluationArtificialIntelligence #AIWorkforceTraining #StateAICitizenTraining  #StatePerformanceMeasurementArtificialIntelligence #StateHRandArtificialIntelligence #StateGovernmentWorkforceandArtificialIntelligence #StateandLocalManagementNews #StateandLocalArtificialIntelligenceNews #BarrettandGreeneInc