LANSING — Michigan officials are exploring the use of artificial intelligence to help detect fraud in the state’s food assistance program, but legal experts warn the technology could wrongly flag people who actually qualify for benefits.
The state is looking at using AI tools to review cases within the Supplemental Nutrition Assistance Program, also known as SNAP, with the goal of improving accuracy and reducing payment errors.
Officials say the system could scan cases more efficiently and identify potential issues before benefits are distributed. However, experts caution that many cases flagged as fraud are actually the result of mistakes, not intentional wrongdoing.
They also point to Michigan’s past issues with automated systems, including the state’s unemployment fraud detection program, which falsely accused thousands of residents and led to years of legal challenges.
More than one million Michiganders rely on SNAP benefits, and experts say any new system must ensure accuracy and accountability to avoid repeating past mistakes.