Uber’s robotaxi uses advanced Level 4 autonomous driving technology with cameras, lidar, and radar to reduce human error, which causes over 90% of accidents. But self-driving cars can still face trouble understanding unpredictable human actions like those seen near construction zones or pedestrian crossings. This creates a legal puzzle: if an autonomous taxi crashes, who is responsible? In normal cases, victims sue negligent human drivers. But a robotaxi has no human driver to blame or question in court. Various countries try to solve this in different ways. Germany holds the vehicle’s owner strictly liable. Each vehicle must have a “black box” to help decide if the automated system caused the crash; the owner can then sue the manufacturer. In the UK, victims claim from the insurer, who can then seek compensation from the manufacturer. The US has mixed state laws, sometimes protecting manufacturers, sometimes holding them liable. These laws aim to make it easier for victims to get compensation. But with private ownership of robotaxis growing, proving fault in the algorithm is tough. Algorithms are secret and complex, making legal proof hard. Should laws remain the same or change to demand manufacturers prove their tech worked right? Or share liability with owners? The legal system faces an "uber-complicated" problem. Martin Wong, honorary secretary of the Hong Kong Bar Association, highlights this challenge in his Legal Tales column.