인공지능과 계약법 - 인공 에이전트에 의한 계약과 사적자치의 원칙 -

The rapid development of A.I. technology impacts greatly on law of contract. Beside the fundamental issues which could arise if so-called ‘strong A.I.’ comes into existence, there are many difficult issues even in the present phase of ‘weak A.I.’ These issues arise from the autonomous feature of A.I. and super-rationality of A.I. in certain areas. These features stand out most in the contract formation cases by artificial agents. When contracts are concluded by artificial agents, the autonomous feature of A.I. could bring out such results that were not intended or even opposed by users of them. This precipitates the need for legal explanation for attributing the liability to the users. There is also the issue of adjustment of conflicting interests of the parties. When legal acts are conducted which are opposed by the users, the resulting risk should be appropriately apportioned between parties. So it should be examined what the adequate criterion of risk apportionment is and what kind of legal princple could be resorted to for this purpose. This paper tried to demonstrate that the good old principle of private autonomy could hold fast as the basis for justification of contracts by artificial intelligence and that appropriate risk apportionment could be achieved by analogy of the apparent agency authority. A.I. features rationality that surpasses that of human in some areas. These phenomena have long been around us, but raise new problem of asymmetry, which is ‘rationality gap’ in the area of law of contract as a result of combination with the autonomous feature of A.I. There is structural asymmetry between people who use advanced A.I. for contract and people who don’t. This prompt the need for the study of legal apparatus for the protection of the counterpart of the A.I. user. This paper attempted to show that the expansion of right of withdrawal and internalization of the principle of good faith into artificial agents could help solve this problem. The problem of autonomous feature of A.I. and rationality gap could be interpreted as challenge to the principle of private autonomy. The development of A.I. technology concerning law of contract raises questions whether the fact that artificial agent user is liable for the result which does not correspond to his intention infringes his self-determination, and whether the self- determination of the other party who transacts with artificial agent user reduces to emptiness. These questions can be rephrased into the question whether the principle of private autonomy can remain valid in the age of A.I. and what kind of complementary measures are needed for its doctrinal power to be maintained. This paper aims at answering properly to these questions.