The Accountability Gap: When Autonomous Machines Fail in Public Spaces

Last weekend, a patrolling humanoid robot fell down a staircase at Hong Kong International Airport, shattering its arms. Fortunately, no one was hurt, and the area was promptly closed off.

But imagine the exact same robot in the public arrivals hall—with a traveller on those stairs. The physics don't change. The outcome does.

This incident brings us to a critical question that every operator, contractor, insurer, and regulator is about to face: When an autonomous machine injures a person... who pays?

Is it the operator? The contractor? The manufacturer? Or the insurer, operating under a policy that was never designed for a machine that makes its own decisions?

The Mismatch in Modern Insurance

Traditional insurance prices risk once a year. An autonomous machine's risk changes many times a second. That mismatch isn't just a small gap; it's an entire uninsured category. And it is arriving at the exact moment physical AI scales from thousands of machines to millions worldwide.

This is the problem YAS is built for. With AURA, we are bridging the gap between dynamic machines and static insurance. AURA turns live machine telemetry into a live risk score. This helps coverage follow the machine, helps insurers adapt pricing to real behaviour, and supports faster claims resolution.

We ensure that insurers keep the authority, while we give them the signal.

The next robot won't fall in a closed test zone. It will fall on a public floor. The real question is whether we have made that world insurable before it arrives.

At YAS, we provide risk intelligence for the machine economy.