近日，北京互联网法院审结了一起侵害作品署名权和信息网络传播权纠纷案件，该案为我国首例涉 AI 文生图著作权案。原告使用开源软件 Stable Diffusion，通过输入提示词的方式生成了涉案图片后发布在小红书平台。被告在百家号上发布文章，文章配图使用了涉案图片。原告认为，被告未经许可使用涉案图片，且截去了原告在小红书平台的署名水印，使得相关用户误认为被告为该作品的作者，严重侵犯了原告享有的署名权及信息网络传播权，要求被告公开赔礼道歉、赔偿经济损失等。
法院经审理认为，从涉案图片的外观上来看，其与通常人们见到的照片、绘画无异，显然属于艺术领域，具有一定的表现形式。涉案图片系原告利用生成式人工智能技术生成的，从原告构思涉案图片起，到最终选定涉案图片止，原告进行了一定的智力投入。因此，涉案图片具备“智力成果”要件。从涉案图片本身来看，体现出了与在先作品存在可以识别的差异性。从涉案图片生成过程来看，原告通过提示词对人物及其呈现方式等画面元素进行了设计，通过参数对画面布局构图等进行了设置，体现了原告的选择和安排。因此，涉案图片具备“独创性”要件。涉案图片是以线条、色彩构成的有审美意义的平面造型艺术作品，属于美术作品，受到著作权法的保护。
根据法院判决，被告未经许可使用涉案图片，侵害了原告的信息网络传播权，应承担侵权责任。此外，被告将涉案图片进行去除署名水印的处理，侵害了原告的署名权，应当承担侵权责任。
此次判决为我国首例涉 AI 文生图著作权案，案件审结表明，AI 生成的图片也受著作权保护，且使用 AI 生成的图片需经过著作权授权，否则构成侵权。
新闻翻译：
Title: China’s first AI-generated image copyright case is ruled in favor of the plaintiff
Keywords: AI-generated image, copyright, court ruling
News content:
Recently, the Beijing Internet Court has ruled in favor of the plaintiff in a case involving copyright infringement of an AI-generated image. This case marks the first time in China that an AI-generated image has been ruled to be protected by copyright.
The plaintiff used the open-source software Stable Diffusion to generate the image and publish it on the Xiaohongshu platform. The defendant posted an article on the Baijiahao platform, using the image as the main picture. The plaintiff argued that the defendant had not obtained permission to use the image and that the image had been misattributed to the defendant, seriously infringing the plaintiff’s rights to authorship and information network distribution.
The court held that the image had the characteristics of an art work and was protected by copyright. The image was generated by the plaintiff using the generative AI technology, and the plaintiff invested a certain amount of intellectual effort in the design of the characters and other visual elements, as well as the setting of the image layout and composition. Therefore, the image qualified for the “intellectual achievement” requirement. Furthermore, the difference between the image and the prior art could be identified. The process of generating the image involved the plaintiff’s input of prompt words and the selection and arrangement of parameters, demonstrating the plaintiff’s choice and judgment. Therefore, the image qualified for the “originality” requirement. The image is a piece of graphic art consisting of lines and colors, and it is protected by copyright law.
The court ruled that the defendant’s use of the image without permission violated the plaintiff’s rights to information network distribution and ordered the defendant to make a public apology and compensate the plaintiff for damages. Additionally, the defendant’s removal of the plaintiff’s watermark on the image violated the plaintiff’s right to authorship, and the defendant should also bear the responsibility for the infringement.
This ruling marks the beginning of the protection of AI-generated images under copyright in China. The case demonstrates that AI-generated images are also protected by copyright and that the use of AI-generated images requires authorization. Otherwise, it constitutes copyright infringement.
【来源】https://www.ithome.com/0/742/017.htm
Views: 1