Everyone hates traffic. Let’s use AI to fix it. 

In a politically polarized country where opinions often clash, one frustration unites us all: traffic. Every year, Americans collectively lose billions of dollars and millions of hours to traffic congestion, with the average commuter wasting 51 hours and over $1,000 annually. Our roads are stressed, our infrastructure is aging and local transportation departments are being asked to do more with less. 

The U.S. has the unfortunate distinction of leading the developed world in total car crashes, racking up hundreds of billions of dollars in damage and delays.  

Meanwhile, the U.S. is leading in AI development and investment. Stanford's 2025 AI Index Report shows U.S. institutions produced 40 notable AI models, compared to China's 15 and Europe's three. Private investment reached $109.1 billion last year — 12 times more than any other leading country. 

It’s time to connect the dots. 

AI can help transportation agencies do what decades of “hard infrastructure” improvements couldn’t: significantly improve transportation challenges that have plagued American cities for generations. By enabling real-time traffic analysis, preventing collisions, optimizing emergency response and detecting hazards, AI directly reduces traffic fatalities through smarter, safer infrastructure.  

Some cities are beginning to embrace this future. In Bellevue, Wash., city officials used AI-powered machine vision to analyze footage from traffic cameras — identifying patterns that lead to crashes and adjusting signal timing and infrastructure accordingly. The result? A 20 percent reduction in crashes at intersections where the system was deployed.   

In Houston, a Smart City program uses AI-powered edge computing to monitor traffic in real-time. It tracks vehicles, detects congestion and alerts officials to hazards. The system can even adjust traffic lights dynamically based on live data — helping to prevent backups before they start.

And the best part? This doesn’t require building from scratch. AI helps optimize the system we already have delivering measurable benefits without busting the budget. Imagine a small city or county trying to build safer streets for people with disabilities, but without the big budgets or resources that larger cities have. AI can dynamically adjust crosswalk timing based on real-time pedestrian detection, accommodating wheelchair users or slower walkers without significantly delaying transportation flow. At a time of constrained federal and state budgets, that’s the kind of return on investment we can’t afford to overlook.  

To make this a reality across all 50 states, Congress should treat digital infrastructure with the same urgency as asphalt and steel. The main legislative vehicle in Congress governing transportation infrastructure is due for renewal in 2026. As lawmakers prepare the next surface transportation bill, they should consider harnessing this AI opportunity in three ways: 

• Incorporate digital infrastructure and AI technologies into existing federal transportation planning, funding and research programs;  
• Create a dedicated digital infrastructure program, as proposed by ITS America, to bring proven traffic tech to more communities; 
• Invest in workforce training to help agencies deploy those tools effectively.  

Beyond the next transportation bill, digital infrastructure should be a key part of the U.S. Department of Transportation Innovation Agenda — because it’s not just about new tech, it’s about better outcomes for all Americans. This isn’t a partisan idea — its common sense. AI can help us get where we’re going faster, safer and smarter. Unlike traffic, that’s something we can all get behind. 

Gerry Petrella is general manager for U.S. public policy at Microsoft.