FameEX Hot Topics | Japanese AI Experts Sound the Alarm on Bots Trained with Copyrighted Material

Japanese lawmaker Takashi Kii is actively leading the charge for regulatory measures aimed at safeguarding copyright holders against potential infringements facilitated by artificial intelligence (AI) algorithms. Concerns have been raised by Japanese AI experts and researchers who fear a surge in copyright infringement cases, along with detrimental consequences such as job losses, dissemination of false information, and unauthorized disclosure of confidential data, due to the use of illegally acquired information to train AI models.

On May 26, the government's AI strategy council submitted a draft expressing apprehension about the lack of regulations surrounding AI, particularly regarding the risks of copyright infringement. Takashi Kii emphasizes the absence of laws in Japan that explicitly prohibit the utilization of copyrighted material and unlawfully obtained information for training AI models.

Takashi stated, "In Japan, regardless of whether it is for nonprofit or for-profit purposes, or for acts other than duplication, information analysis by AI heavily relies on data obtained from illegal sources." He also raised concerns about the absence of guidelines governing the use of AI chatbots like ChatGPT in schools. This issue becomes even more pertinent considering the potential adoption of this technology in the education system as early as March 2024.

Minister Nagaoka provided an inconclusive response, stating that implementation would occur "as soon as possible" without specifying a definite timeline. Andrew Petale, a lawyer and trademarks attorney at Y Intellectual Property in Melbourne, explained that the subject remains in a "gray area."
Petale elaborated, "Copyright protects the expression of ideas rather than the ideas themselves. In the case of AI, a human being inputs information into a program." He further clarified, "The inputs come from people, but the actual expression is generated or produced by the AI itself. Once the information is inputted, it essentially becomes beyond the control of the person, as the AI generates or produces the output."

Recognizing machines or robots as capable of authorship is a crucial aspect that must be addressed in legislation. Currently, this realm remains ambiguous and uncharted, necessitating legal proceedings and regulations to resolve various hypothetical scenarios. Petale underscored the importance of determining whether the responsibility lies with AI creators for developing tools used for copyright infringement or with the individuals employing those tools.

AI companies generally argue that their models do not infringe on copyright, as they transform original works into something new, which qualifies under the fair use provisions of U.S. laws, where most of the legal discussions are taking place. Resolving these issues requires a comprehensive approach involving lawmakers, legal experts, AI companies, and copyright holders to strike a delicate balance between protecting intellectual property rights and fostering the development and responsible use of AI technology.

Disclaimer: The information provided in this section is for informational purposes only, doesn't represent any investment advice or FameEX's official view.