英文摘要 |
In order to put an automatic driving vehicle (AV) into practicaluse, it is necessary to reduce the probability of unavoidable accidentsto a socially acceptable level. That level is the socially acceptablelimit for 'accidents that are not attributed to anyone.'Next, regarding the criminal liability of AV (the AI that controlsit) itself, as long as the essence of 'criminal liability' is valid for thebehavior of AI, it should be affirmed. However, if AI has standardnormative and moral capabilities, there is no 'criminal liability' forthe accidents it causes. In addition, if AI can be held criminallyliable, it requires equality and equality with natural persons in thelaw.In addition, it is necessary to consider the legal responsibilitiesof persons involved in autonomous driving separately for 'driver'(occupant), automobile owners, automobile manufacturers, controlprogram developers, data providers, and national and local governments.Furthermore, for emergency response programs that save a largenumber of people at the expense of a small number of people, such asthe 'trolley problem,' the norms of society themselves need to bedeepened and developed. In addition, it is necessary to solve the'self-sacrifice program' that involves the sacrifice of occupantsbased on the principle of emergency evacuation.Finally, in the spread of autonomous driving, the future of thedriver's license system, the choice of public transportation or privatecars, the review of traffic laws, and the avoidance of criminalliability by the accident investigation system should be considered. |