人工智慧創作品之著作權保護──從繪畫機器人談起

Abstract

本文首先透過繪畫機器人之介紹，使讀者理解機器人獨立自主從事創作之情境，並進而激發問題意識：當機器人已具備某種感知能力，能夠自行生成人類所無法預知內容的作品（AI-generated work，以下簡稱「人工智慧創作品」）時，著作權相關規範應如何調整以因應此種新興作品之保護？為回應此一議題，本文主要分為法律與政策兩方面探討：在法律方面，參考英國立法例，初步建議可於著作權相關規範中將人工智慧創作品納入保護，其具體內容考量法律安定性，則考慮透過擬制既有制度以填補目前法制空缺，經檢視現行著作權法規範中與人工智慧創作品之創作邏輯最為接近者，發現聘雇著作的法理與人工智慧創作品有一定相似性，其中又以職務著作最為接近，故似可以此作為人工智慧創作品擬制之依據；另在政策方面，為求人工智慧創作品之長期且全面性發展，初步建議可參考日本之相關討論，設立跨部會整合之專責機構加以統籌擘劃。本文結論強調，賦予此一新型態創作品法律位階的保護，且以人類為權利主體，始能達到創造誘因、激勵創新之目的，才更有機會建立一個正向循環的人工智慧經濟，並期待政策上將人工智慧創作品納入人工智慧產業之整體規劃，以使產業全面發展。

Starting with newly-invented painting robots, this article hopes to inspirereaders’ awareness of current issue: how to modify copyright framework, when therobot is perceptive and capable of generating unpredictable artwork? This articletries to respond this issue from two perspectives of law and policy. Firstly in legalperspective, this article refers to UK’s similar mechanism, claiming that protectionneeds to be built in the copyright related laws. As for the protection method, basedon the principle of legal stability, analogizing AI-generated work to the existingwork under copyright law may be a feasibly way. After reviewing each type of theworks, the author considers work-made-for-hire theory has the highest similaritywith AI-generated work. Secondly in policy perspective, this article refers to EuropeanUnion and Japan’s policy consideration to introduce the advanced mechanismfor Taiwan’s reference. The initial suggestion is proposed that an inter- departmental sector shall be established for the long-term development of the AIgeneratedwork. In conclusion, the author emphasizes that legal protection shouldbe applied to this new type of work and attribute right to human, so that the incentivefor innovation will be created and eventually benefit to Taiwan’s AI industry.Also, the policy on AI-generated work should be formulated, so that the Taiwan’sAI development blueprint will be more complete.