In Japan, discussions on the institutional design governing intellectual property in relation to generative AI have finally moved into full swing. A symbolic development in this process is the opinion paper published by the Japan Newspaper Publishers & Editors Association (JNPEA) regarding the government’s proposed rules on intellectual property protection for generative AI operators.
An Evaluation That Signals Both “Progress” and “Concern”
The government’s proposal, referred to as the “Principles Code (tentative name),” calls on generative AI operators to disclose what measures they are taking to protect intellectual property. The JNPEA evaluated this as “important content” that advances the appropriate protection of news content and similar materials.
At the same time, however, the opinion paper strongly emphasized concerns about the proposal’s effectiveness. Because disclosure is voluntary and not accompanied by penalties or enforcement mechanisms, it points out that “whether operators will actually comply remains unclear.” In other words, while the principles themselves may be sound, whether they will function in practice is an entirely different matter—a sober and realistic assessment.
The Limits of Principles-Based Rules
The proposed rules fall into the category of so-called “soft law.” In rapidly evolving technological fields, such an approach certainly has the advantage of flexibility. However, the relationship between generative AI and news content is no longer a “future concern”; it is a business issue unfolding in real time.
For news organizations in particular, the unauthorized training and use of their content directly affects their revenue foundations and editorial independence. Under such circumstances, how much deterrent effect can a framework have if compliance is left entirely to the voluntary discretion of operators? The doubts raised by the JNPEA are highly practical and grounded in reality.
From “Requests for Compliance” to “Legislation”
In its opinion paper, the JNPEA urges the government to actively encourage compliance among operators that do not follow the rules. It goes further, recommending that if no improvement is seen, the government should promptly consider formal legislation.
For the Cabinet Office, the recipient of the opinion, this is a weighty message. It suggests that the phase of “waiting to see how guidelines work” cannot be prolonged much longer.
A Turning Point Between Innovation and IP Protection
Preventing obstacles to the development of generative AI while appropriately protecting existing intellectual property are not mutually exclusive goals; they are challenges that must be reconciled. This opinion paper should not be read merely as an industry request, but rather as a warning that “principles alone cannot maintain order.”
Japan’s generative AI policy now stands at a critical crossroads. Will the Principles Code remain an empty promise, or will it evolve into a set of rules with real effectiveness? The government’s decisions going forward are likely to have a profound impact on the future relationship between generative AI and journalism.