The Exploitation of Self-Driving Vehicles

While discussing autonomous systems with researchers working in robotics, I came across a term that remained with me: "exploitation of self-driving vehicles." It stood out because it reframes a set of common interactions that we don't often describe this way.

We often discuss trust in self-driving vehicles, but less about how people adapt their behavior around them, especially with robotaxis. As AVs become more predictable, socially compliant, and safety-focused, humans respond. There are few instances of humans willingly stopping AVs by wearing t-shirts which contain traffic symbols.

This is not simply a perception or sensing issue, but also a social value question.

One helpful lens is Social Value Orientation (SVO), which describes whether an agent acts in a pro-social way, prioritizing collective good, or an egoistic way, prioritizing self-interest. Many autonomous vehicle behaviors today are intentionally pro-social by design, focused on safety, caution, and de-escalation.

The interaction between these systems and human behavior is interesting, as it does not always mirror the same orientation. As robotaxis enter real city settings, this seems like an important area to explore further. It adds one more layer to how we think about trust in autonomous mobility.