Artificial intelligence is becoming part of everyday mobility. While often linked to autonomous vehicles, its impact on road safety already extends far beyond self‑driving cars. From detecting driver distraction to improving emergency response, AI technologies are helping make roads safer, traffic more efficient and mobility more intelligent.
Recognising this potential, the International Telecommunication Union (ITU), together with the UN Secretary‑General’s Special Envoy for Road Safety and the Office of the UN Envoy on Technology, launched the AI for road safety initiative in 2021. It promotes an AI‑enhanced Safe System approach aimed at reducing global traffic fatalities.
It focuses on six key pillars:
- road safety management
- safer roads and mobility
- safer vehicles
- safer road users
- post‑crash response
- speed management
Europe is also investing heavily in AI‑driven innovation. The IVORY project is developing scalable AI technologies for proactive infrastructure safety management and user support through AI‑based vehicle automation.
Smarter roads through better data
Road crashes are shaped by many factors, including driver behaviour, traffic conditions, weather and infrastructure quality. AI helps clarify how these elements influence crash risk by gathering and organising data from crash reports, smartphones, in‑vehicle sensors and public cameras. By analysing this information proactively, AI enables authorities to act before safety is compromised.
Nudging road users to safer behaviour
AI’s strength lies in processing large volumes of information quickly and spotting patterns humans miss. This enables faster, smarter and more proactive safety interventions.
Driver monitoring systems
AI tracks facial expressions, gaze, posture and hand movements to detect distraction, drowsiness or impairment. When unsafe behaviour appears, it issues immediate warnings or activates corrective functions.
Advanced driver assistance systems
Many vehicles now rely on ADAS, which read the road, monitor driver attention and interpret traffic conditions in real time. They can alert the driver or automatically intervene, from braking to steering adjustments, to prevent dangerous situations.
Detecting dangerous traffic situations
AI can identify wrong‑way driving, unusual traffic patterns and other early signs of danger. By analysing near‑misses and video‑based behaviour, these systems pinpoint high‑risk locations before crashes occur, helping authorities act earlier and more precisely.
Improving emergency response
AI also plays a critical role after a crash. Faster emergency response can significantly reduce injury severity and save lives.
AI is increasingly used to:
- detect incidents more quickly
- inform emergency responders automatically
- route ambulances efficiently
- support emergency call handling
By improving response times and situational awareness, AI strengthens post‑crash care and emergency interventions.
The importance of responsible AI
AI’s potential depends on access to high‑quality data and on using it responsibly. This remains one of the biggest challenges in road safety.
Data often sit in isolated systems or “silos”, limited by privacy concerns and fears of legal exposure. AI models also require skilled professionals to validate outputs and ensure reliable performance.
Transparency, accountability and responsible data governance are essential. Stakeholders must promote trustworthy AI practices that deliver safe, fair and socially beneficial outcomes.
Inclusion is equally important. Underrepresented groups must be visible in road safety datasets to avoid biased results. AI systems should support safer mobility for all road users.
Towards a safer future
Artificial intelligence is rapidly reshaping road safety. From preventing crashes and improving driver behaviour to supporting emergency services and helping authorities understand risk, AI offers powerful tools to reduce injuries and fatalities.
But technology alone is not enough. Success will depend on responsible data governance, strong collaboration and a continued focus on human‑centred safety.
As AI evolves, it has the potential not only to make transport smarter and more efficient but ultimately to save lives.
Want to discover how our members implement AI in road safety?
- Improving incident detection through real-time data analytics and AI
- Improved road safety with AI inspections
Alongside these examples, our Good Practice of the Month for August highlights an innovative contribution from the Central Traffic Management Office of the Free State of Bavaria. Their Traffic Light of the Future Initiative uses AI‑enabled detection and intelligent signal control to improve safety at intersections where pedestrians and cyclists face the greatest risks. By combining camera and radar sensors with real‑time decision‑making, the system identifies vulnerable road users and potential conflicts instantly, enabling adaptive green phases, targeted warnings and safer passage for emergency vehicles. Early testing shows near-perfect detection accuracy and clear safety gains, demonstrating how intelligent infrastructure can reduce turning collisions and create more predictable environments for all road users. With strong performance and clear potential for replication, the initiative shows how AI can deliver practical improvements to everyday mobility.
Have a best practice or success story to share in this area? If you are an existing member and signed in, then submit them here.
Sources:
Artificial Intelligence in Proactive Road Infrastructure Safety Management. (2021).
Tselentis, D. I., Papadimitriou, E., & van Gelder, P. (2023). The usefulness of artificial intelligence for safety assessment of different transport modes. Accident Analysis & Prevention, 186, 107034. https://doi.org/10.1016/j.aap.2023.107034
Yannis, G., & Ziakopoulous, A. (2024). Artificial Intelligence for Road Safety.
