Key takeaways:
AI is here to stay in transportation systems and three research institutes are coming together to determine best uses and practices.
The AI tools available to transportation agencies today didn't exist two years ago, and many will be outdated two years from now.
Collaborating institutions, including University of Michigan Engineering, University of California Berkeley and Texas A&M.
Three of the nation's leading transportation research institutes have launched an effort to help public transportation agencies use AI effectively as the technology rapidly advances.
The National AI Consortium for Transportation , or NAICT, brings together the University of Michigan Transportation Research Institute , the University of California Berkeley Institute of Transportation Studies and the Texas A&M Transportation Institute into a national research initiative to give agencies a credible, noncommercial resource for evaluating, deploying and governing AI.
The AI tools available to transportation agencies today didn't exist two years ago, and many will be outdated two years from now. New models, products and vendor promises arrive every month, each claiming to make roads safer, freight move faster or infrastructure last longer. For the public agencies responsible for those systems, the hardest question is no longer whether to use AI. It's which tool fits which problem, how to use it well, and how to tell a proven solution from a polished sales pitch.
An inclusive approach
NAICT is not being built behind closed doors. The consortium is inviting state and local transportation agencies, industry, professional organizations and researchers to get involved now, starting with listening sessions this fall and a panel and workshop at the Transportation Research Board Annual Meeting in January 2027.
"NAICT is designed to give agencies and industry a single, credible, noncommercial resource for evaluating, deploying and governing AI in real-world transportation operations," said UMTRI Director Henry Liu. "We need to have a guided discussion to help shape what AI is and what it can do in safe, responsible and effective ways."
Liu co-leads NAICT with Scott Moura, director of Partners for Advanced Transportation Technology, an ITS Berkeley research center, and Greg Winfree, director of the Texas A&M Transportation Institute.
What's next?
NAICT will align its priorities with agencies and practitioners. Initial activities include:
UC Berkeley Listening Session (Nov. 5, 2026) : NAICT will hold a workshop and listening session. Transportation leaders and practitioners will have an opportunity to share the AI challenges they face and identify resources and support needed to move forward.
Transportation Research Board Annual Meeting (January 2027): NAICT will host a panel and interactive workshop in Washington, D.C., bringing together experts from research, industry and policy to explore safe and effective AI deployment.
Education and Governance Resources (early 2027): NAICT plans to introduce its first educational offerings and a draft AI governance framework. Agencies and partners who engage early will have opportunities to review and help shape these resources before their broader release.
"We're moving fast, and we'd rather build this with agencies than for them," Liu said. "If you're responsible for putting AI to work on public roads, we want you in the room. Come find us at TRB."