U.S. — Lyft updated its AV Partner Safety Evaluation Framework on June 22, 2026. The revised framework requires autonomous vehicles operating on its platform to implement a multi-modal redundant perception system with sensor diversity.

This policy applies to fully driverless vehicles operating on the Lyft platform. It does not apply to driver-assistance features used by human drivers on the platform. David Risher, CEO of Lyft, addressed the rationale for the new requirement. "To meet our safety standards, autonomous vehicles must have a multi-sensor approach before we allow them on the Lyft platform," Risher said.

Lyft may revisit its multi-sensor policy under specific conditions. These include if safety data shows single-sensor systems perform comparably to established multi-modal architectures. Additionally, the policy could be re-evaluated if a single-sensor AV system meets a standard for fully driverless operations set by a credible authority such as the National Highway Traffic Safety Administration (NHTSA).

Multi-modal perception systems incorporate various sensor types to overcome the limitations of individual sensors. Cameras used in autonomous vehicles can be blinded by glare, fog, and lens occlusion. Radar systems can struggle with detecting stationary objects. LiDAR technology can experience degraded performance during heavy precipitation. "AVs never drive distracted and are designed to obey speed limits and traffic laws," Risher said.

Approximately 40,000 deaths and 2.4 million injuries occur on U.S. roads each year. Car crashes represent the leading cause of death for teens and young adults in the U.S. Speeding, alcohol impairment, and distracted driving are identified as the largest contributors to road fatalities. "Autonomous Vehicles (AVs) are part of that plan," Risher said.