AI in transit: the foundation is not the algorithm

AI won't transform transit on its own. Success depends on trusted data, governance and better operational decisions.

In brief

• AI in transit depends on trusted data, operational context and governance before advanced tools can deliver value.
• Transit modernization starts by breaking down information silos and focusing on practical, outcome-driven use cases.
• Responsible AI can help agencies improve reliability, strengthen public trust and support better decisions over time.

AI is not the starting point for transit operations looking to modernize. It’s the remarkable capability that becomes possible when agencies get the right data foundation in place.

This comes up in so many conversations we’re having with clients in Canada and beyond. Stakeholders in cities around the world are excited about using AI to make transit better, faster, easier and cost effective. I am, too. But with every discussion of the possibilities, I come back to the same core truth: success is not about the algorithm. It starts with the right foundation.

AI can be a huge help. First, transit must reinforce the trusted data, operational context and human judgement needed to make AI useful, safe and credible. With credibility comes the confidence citizens need to trust in the transit system and embrace its potential well beyond the heady days of World Cup fever and other major events.

To move in this direction, transit leaders in Toronto can take a progressive approach, doubling down on incremental priorities:

1. Put the groundwork in place now. AI relies on a solid, connected set of good data. Think beyond data migration exercises and begin dismantling the operational silos that fragment information. Transit agencies must know what data they have and what it means before moving into more advanced AI tools. Remember: the data doesn’t need to be perfect. The greatest value comes from starting with available data, proving outcomes through targeted use cases and improving data quality over time.
   
   
2. Evaluate AI opportunities based on practical operational value — not novelty. To make more of AI, transit agencies must think beyond the hype. Start with lower-risk use cases where AI can support disruption response, predictive maintenance or operational readiness. Prioritize trusted data and digital twins before scaling AI across the entire network. Pilot, learn, measure, adapt as you go, and build for production, even if you’re just designing for a pilot.
   
   
3. Focus on responsible AI governance by design. In 2026, the EY AI Sentiment Survey showed 71% of Canadians are concerned with AI systems being hacked or breached. Two-thirds worry that organizations won’t be accountable for AI driven harm. Transit agencies must proactively mitigate that fear to build trust in the systems and confidence in the transportation network. That means focusing on responsible AI adoption grounded in principles like accountability, data protection, reliability, security, transparency, explainability, fairness, compliance and sustainability.
   
   
4. Treat AI as decision support for safer, more reliable transit operations. Whether deploying digital twins or any other form of AI, transit agencies cannot afford to frame use purely as automation. We need to shift our thinking away from the idea that AI can transform transit on its own. This is about using AI to support better transit decisions grounded in trusted data, digital twins, operational context, governance and human insight.

Soccer brought the world to town. And transit in Toronto and Vancouver rose to meet the moment. By building on those lessons and using AI responsibly, transit leaders across Canada can strengthen service, earn public trust, supporting broader economic and social progress.

Summary

AI represents a significant opportunity for transit modernization, but successful adoption begins with trusted data, connected systems and responsible governance. By focusing on practical use cases, strong operational foundations and human-centred decision-making, transit agencies can use AI to improve reliability, safety and rider experience while strengthening public confidence in the transportation network.