State DMVs are running AI systems right now. Brian Shell, a senior partner at Kyndryl, details in a recent interview how government agencies are deploying AI to cut wait times and streamline services that Americans interact with daily. This is not a pilot program or a roadmap. It is operational.
The piece matters because government AI adoption is the story most outlets miss while chasing OpenAI announcements. Shell walks through specific use cases at the DMV level, which means the infrastructure, the constraints, and the tradeoffs are concrete. The source is a vendor with skin in the game, so the original is worth reading with that lens applied.
What comes next is the real question the article raises: if legacy agencies like DMVs can deploy AI at scale, the ceiling on public-sector adoption is higher than critics assume. Read the full interview for the specific applications Shell names and the implementation details that explain why this is harder than it sounds.
[READ ORIGINAL →]