In unavoidable traffic accidents, autonomous vehicles (AVs) face the dilemma of protecting either the passenger(s) or third parties. Recent studies show that, in accidents involving an AV, most people prefer the AV to behave such that it minimizes expected harm. At the same time, however, they are more willing to adopt an AV if it prioritizes the passenger(s), i.e. themselves. A regulator therefore needs to set a rather high level of driver protection in order to increase the market penetration of AVs. Using a game-theoretic model, we analyze how the second-best optimal level of AV passenger protection depends on (i) the AV safety advantage, (ii) the intensity of drivers’ other-regarding preferences, and (iii) the (preference) cost of AV adoption. We show that a higher AV safety advantage may either increase or decrease the second best optimal level of AV driver/passenger protection. In an extension, we find that increasing the degree of driver protection has the additional benefit of leading to higher care levels of other road users such as pedestrians and cyclists.

中文翻译：

---

自动驾驶汽车：道德困境和采用激励措施

在不可避免的交通事故中，自动驾驶汽车（AV）面临着保护乘客还是第三方的困境。最近的研究表明，在涉及自动驾驶汽车的事故中，大多数人都希望自动驾驶汽车能够最大限度地减少预期伤害。然而，与此同时，如果自动驾驶汽车优先考虑乘客（即他们自己），他们更愿意采用自动驾驶汽车。因此，监管机构需要设置相当高水平的驾驶员保护，以提高自动驾驶汽车的市场渗透率。使用博弈论模型，我们分析了自动驾驶汽车乘客保护的次优水平如何取决于（i）自动驾驶汽车安全优势，（ii）驾驶员其他相关偏好的强度，以及（iii）（偏好） ) AV 采用成本。我们表明，更高的自动驾驶安全优势可能会增加或减少自动驾驶驾驶员/乘客保护的第二最佳水平。此外，我们发现，提高驾驶员保护程度还有一个额外的好处，即可以提高对行人和骑自行车者等其他道路使用者的护理水平。