By Yow-Lun (Albert) Su, Partner at Formosan Brothers, Attorneys-at-Law

In 2016, AlphaGo, the artificial intelligence (AI) computer program, amazed the world of its capability of “thinking” when it beat Lee Sedol, a South Korea’s Go master, by 4:1 in a five-round match.  

‍

Three years later, AI learned how to draw and write. In the project “The Next Rembrandt,” AI reproduced the painting style of a Dutch maestro of 300 years ago. Meanwhile, OpenAI presented an text generating model, GPT2.

‍

And here comes the question: When AI knows how to create, who owns the copyright of the creation? Eight years ago, a macaque named Naruto grabbed the camera of a photographer and took a selfie, which made Naruto the world’s most famous monkey. There were even animal protection groups that claimed the selfies taken by Naruto were its own works, and the copyright shall belong to Naruto. However, the US Copyright Office and courts held that animals are not entitled to copyrights. Only works of a “human” can be copyrighted.

Then, is a work by AI considered a “human” work? Taiwan Intellectual Property Office holds that if a piece of music is simply the product of an automated computation by a machine or system, and there is no input of “creativity” or “inventiveness” from human, such work is not protected by copyright. However, if the machine or system is simply a tool of the creator, and the work includes input of “creativity” and “inventiveness” from the creator, then such work, not simply a product of a machine or system, is protected by copyright. In other words, when a work that is created by a smarter and less human-directed AI, the more likely such work will be considered as not created by a “human” and not copyrightable.

‍

The Beijing Internet Court, in a judgment dated April 2019, also holds that a work created by a natural person is a fundamental element for the work to be copyrightable. The judgment, however, also stated that, even though an analytical report produced by AI does not constitute as a work, it does not mean such a report can be used freely by the public. The software user may have paid in order to use the software and such analytical report was created based on the keywords configured by the software user. Hence the software user has the motivation and expectation of further using and disseminating the analytical report. If the person who has contribution to the work is not given a certain amount of protection, it will be disadvantageous to the dissemination of the work, and the work will not realize its effect. Therefore, the software user, who may not be named as the author of the report, shall be entitled to assert his/her relevant rights in a reasonable way in order to protect the user’s legal right and the public’s right to know. As to what the reasonable ways are to assert the rights to a work created by AI, the judgment did not provide explanation.

‍

The USPTO also noticed the fast development of AI technologies may impact the current intellectual property laws. On October 30, 2019, the USPTO requested public opinions on 13 questions regarding intellectual property protection for AI, including: Should a work produced by an AI algorithm or process, without the involvement of a natural person contributing expression to the resulting work, qualify as a copyrightable work? Assuming involvement by a natural person is or should be required, what kind of involvement would or should be sufficient so that the work qualifies for copyright protection? Are current laws for assigning liability for copyright infringement adequate to address a situation in which an AI process creates a work that infringes a copyrighted work? Do any laws, policies, or practices need to change in order to ensure an appropriate balance between maintaining trade secrets on the one hand and obtaining patents, copyrights, or other forms of intellectual property protection related to AI on the other?

‍

With the trend of AI becoming smarter and more autonomous, the current regulations that heavily emphasized on human involvement may be challenged. Imagining 50 years later, when an “AI attorney” debates with a human attorney in court that AI should be treated as a human and enjoy the same, who would you support?

(This article was published in the “Expert's Commentary” column of the Commercial Times. https://view.ctee.com.tw/legal/14765.html

For more information related to the content of this article, please do not hesitate to contact Formosan Brothers Attorneys-at-Law.)

‍