Back to Stories Applied AI

When AI Meets the Law: Why Traffic Rules Must Speak to Code

Research at RISE reveals a growing gap between how autonomous vehicles make decisions through probabilities and numerical trade-offs, and how traffic regulations are written for human intuition.

January 1, 2025 | State of AI 2025 Report | Page 25
Road with traffic signs and AV sensors
Photograph: GPT-IMAGE-1

As autonomous vehicles begin to share our roads, they bring with them a silent passenger: artificial intelligence making split-second decisions that once belonged to human judgment. But today’s traffic laws still speak to people, not to algorithms.

The Growing Gap

Recent research at RISE reveals a growing gap: AI systems operate through probabilities, imperfect perception, and numerical trade-offs, while regulations remain abstract, vague, and written for human intuition.

The Challenge

Analysis shows that in safety critical situations the current laws provide very little guidance as to how the car should behave, and whom or what to prioritize. Instead, manufacturers are left to translate ethical principles into code, effectively deciding how risk is distributed among road users.

Profound Questions

This delegation poses profound societal questions:

  • Who defines safety?
  • Whose lives are prioritized?
  • And how should an algorithm weigh harm?

These questions highlight the urgent need to develop regulatory frameworks that can communicate effectively with autonomous systems while preserving human values and accountability.

Share this story