人工智能生成技術方案的可專利性及其制度因應

由人工智能自主生成的技術方案作為一類全新的發明形式，在理論基礎、權利主體、利益分配等多方面對傳統專利制度提出了嚴峻挑戰。由於人工智能目前已經能夠生成符合可專利性的技術成果，既有專利制度面臨變革。在權利歸屬的制度安排上，需要調整多方參與主體的利益關係，構建基於“二元主體結構”的權利主體規則，以使用者作為基本的專利權主體，兼顧對投資者利益的保護，以防止人工智能算法控制者的壟斷;同時，引入約定優先和公平報酬原則作為合理補充，以實現人工智能生成技術方案專利保護的利益平衡。在權利保護的制度構建上，需要完善人工智能生成技術方案的專利申請授權制度，明確其權利行使的基本原則與規範設計，以此進一步實現個人利益與公共利益之間的平衡，推動人工智能行業的創新發展與技術進步。

As a brand-new invention form, the technology solutions generated by artificial intelligence pose challenges to the traditional patent system in many aspects, such as the right holders, theoretical basis and economic paradigm. Artificial intelligence is now capable of generating patentable technology solutions, so the existing patent system needs changes. In terms of the institutional arrangement of right attribution, we should adjust the interest relations of various participants and construct the right holder rules based on ''the dual-subj ect structure'', taking the user as the basic patent right holder and taking into account the protection of the interests of investors to prevent the monopoly by artificial intelligence algorithm controllers. In terms of the system construction of rights protection, the patent authorization system should be improved and the basic principles and standards designed for the exercise of rights should be clarified, so as to achieve the balance between private interests and public interests, and to promote the innovative development and technological progress of the artificial intelligence industry.