LANSING — A traffic light that turns red or green depending on traffic flow rather than on a preprogrammed schedule.
A system that scans signatures on absentee ballots and allows you to mark signatures you don't think are genuine.
A chatbot that can respond to residents' questions at 3 a.m., when most state employees are asleep.
These are all potential uses for artificial intelligence within state government, and more recently for “generative AI,” which can generate text, images, code, and other output, and in some cases perform forms of learning and problem-solving. . . The technology is generating both excitement about its possibilities and concern about issues such as privacy protection and the potential impact on state employees.
Michigan's Office of Technology, Management and Budget wants to spend about $10 million to study how best to leverage AI to better serve Michigan residents. The proposal includes pilot projects and other studies and is part of Gov. Gretchen Whitmer's 2025 state budget proposal.
“The opportunities are endless,” said Laura Clark, who oversees national technology as DTMB's chief information officer. “But we also must have the utmost focus on protecting our national systems and the national data that we hold.”
Rather than being on the “cutting edge” of AI use in state government, Clark said Michigan wants to “be on the cutting edge of learning,” with pilot projects that do not put personal data or broader state systems at risk. He said he wanted to proceed with caution. .
Today, for example, state employees are allowed to use freely available AI systems such as ChatGPT to generate reports if they wish, even though all the information used in such reports is already publicly available. Only if it's in your domain. Said.
Simpler forms of AI technology, from spell-checking your computer to email programs that suggest possible responses to messages you receive, to blocking suspicious websites for virus scans and cybersecurity, to data analysis tools. has been used by state governments for decades.
Some recent applications, such as MiDAS (Michigan Integrated Data Automation System), used to automatically detect unemployment insurance fraud cases, have been disastrous, leaving tens of thousands of Michigan residents accused of fraud. I did. The state recently settled one lawsuit related to that fiasco for $20 million. Michigan State Police's use of technology available for “predictive policing” is also controversial.
Clark said the exploration of future uses for AI is at a very early stage and he could not comment on whether it would impact the future size of the state's workforce, which currently numbers approximately 50,000.
Ray Holman, a spokesman for United Auto Workers Local 6000, which represents about 16,000 state employees, said workers are always concerned when states consider using new technology, but it's not about jobs. He said it's not just because of the potential for loss. Holman said the state has previously implemented an expensive system that is difficult and time-consuming for state employees to use, and has not been particularly helpful to the people it serves. He cited a glitch-filled child welfare software system called MiSACWIS (Michigan Statewide Automated Child Welfare Information System) as an example.
“This should be discussed more with the unions,” Holman said. “They need to work with us.”
After consulting with state agencies, Clark said the areas where there is the most interest in using AI today include chatbot technology, data analysis, document and video processing, and coding tasks. If not, older IT systems can be modernized to extend their lifespan, he said. Replaced.
Michigan's government already uses chatbot technology on the state's website, michigan.gov, and recently requested solutions from vendors to enhance and expand its use of the technology. The technology could likely answer questions or perform other tasks upon request from residents. Clark said translation services will also be included.
“Multiple chatbots are currently in use across certain michigan.gov websites,” DTMB said in a request letter sent to vendors. “Multiple products are in use with varying levels of success depending on the chatbot's technical capabilities and quality of curation. By moving to a single michigan.gov chatbot implementation, we are all Ensures consistency across websites.
As technology advances, so do the threats, Clark said. For example, hackers attempting to maliciously infiltrate state systems using “phishing” emails sent to state employees are using AI to create phishing emails that are more realistic and easier to fool. he said.
Many states and counties outside of Michigan are using AI technology as a tool to verify signatures on absentee ballots, according to Reuters.
Michigan's Department of State “continues to track the various benefits and risks that AI brings to our operations, but at this time has no plans to pursue AI-based technologies,” spokeswoman Cheri Hardmon said. .
The Michigan Department of Transportation has at least three AI pilot projects underway, spokesman Jeff Cranson said.
One project funded by a federal grant uses special cameras and other technology to alert bus drivers and other equipped vehicles to potential dangers such as pedestrians and bicyclists, Cranson said. He plans to issue a warning.
Another system will use flashing road signs to warn drivers going the wrong way on ramps to Interstate 75, Interstate 375 and Interstate 696, officials said.
A third project, also using cameras, is aimed at improving worker safety in highway construction zones, Cranson said.
Contact Paul Egan: 517-372-8660 or pegan@freepress.com. Follow him on X (@paulegan4).
