私法自治下人工智慧作為民事權利主體之商榷

Discussion on Artificial Intelligence as Subject of Civil Capacity Under Private Law Autonomy

倘若未來的人工智慧發展成熟後，其思考決策能力近似或等同於人類，以此等高階人工智慧作為靈魂的機器人，是不是可以自己的名義為民事法律行為？本文認為，高階人工智慧機器人在滿足下列兩項基本要件時，得作為財產法及契約法上之民事主體：有近似人類思考而自我決定之能力，可未在人類特別或明顯監督下，獨立完成法律上或事實上行為；一定數額之獨立財產或金錢。高階人工智慧機器人之私法自治空間應較自然人緊縮，其所為之法律行為類型及價值宜限制之。此外，承認高階人工智慧機器人為民事主體，不代表利害關係人可以免責；而未來宜以特別法規範高階人工智慧機器人的民事法律關係。

In the future, if artificial intelligence matures, its thinking and decision-making capabilities are similar to or equivalent to those of humans, can robots with such high-level artificial intelligence as their souls be able to conduct civil legal acts in their own names? This article recommends that high-level artificial intelligence robots could be the ones with legal capacities under the property laws and contract laws if at least two fundamental requirements are fulfilled: (1) they own the abilities which are so similar to thinking and self-determination abilities of human beings that they could independently conclude legal or factual acts without significant or obvious supervision of human beings; (2) they own the certain independent value of properties or amount of money. The space of private law autonomy of smart ports should be narrower than the one of nature person. The categories and value of legal acts of the above-mentioned robots could be restricted. In addition, recognizing high-level artificial intelligence robots as subject of civil capacity does not mean that parties interested can be exempted from liability. In the future, special laws should be used to regulate the civil legal relations of high-level artificial intelligence robots.