Photo/Illutration Sam Altman, the CEO of OpenAI, which developed the ChatGPT service, speaks with reporters in Tokyo on April 10, 2023, after his meeting with Prime Minister Fumio Kishida. (Koichi Ueda)

The rapid spread of generative artificial intelligence poses a wide spectrum of risks to society, and the government is working to develop effective strategies to respond to these complicated challenges.

One of these risks is the potential for copyright infringement.

AI learns from the vast reams of information that flood the internet and churns out new creations at speeds beyond human capabilities.

A vortex of vague concern rages among policymakers and regulators, as well as authors, creators and other copyright holders, that unfair or illegal use of creative works and other materials could become widespread.

To address this issue, the Agency for Cultural Affairs, through its Copyright Subdivision of the Culture Council, provided partial guidelines last month on situations that might raise Copyright Law issues.

However, many gray areas remain. The fact the agency received 25,000 comments from the public about the guidelines attests to the widespread anxiety in society over the matter.

One regulatory issue concerns the ability of AI to generate speech using a specific person’s voice. This is because current copyright and other intellectual property laws do not provide sufficient protection for “voices” provided by voice actors.

In addition to copyright infringement, privacy violations, and fake information are among the risks linked to generative AI.

Insufficient sharing of information regarding how AI works and how businesses intend to use the technology seems to lie at the root of these concerns.

The public has no way to find out, for instance, what consideration AI companies are giving to people’s rights and safety. Or what kind of data AI is learning from and how it is being used to generate new output.

Unless AI tech firms appropriately disclose such information, rights holders and users will be unable to engage in dialogue and negotiations with them on an equal footing. Efforts are needed to promote mutual understanding.

Still, concerns persist as draft guidelines for AI businesses worked out by the Ministry of Internal Affairs and Communications and the Ministry of Economy, Trade and Industry only call for “voluntary” initiatives by companies to ensure AI’s safety and transparency.

Regarding measures against online defamation, foreign AI-driven social media providers are failing to make sufficient effort to tackle this serious problem despite their overwhelming influence and dominant presence in cyberspace.

Since influential AI powerhouses are also based overseas, how to properly regulate them poses a challenging question for Japanese regulators and policymakers.

Would voluntary approaches work effectively? Is there a need to consider more proactive measures? Taking into account international trends, the government needs to engage in discussions that directly address these questions.

History shows that with the emergence of new transformative technologies, humanity has always gone through a process of trial and error.

Only a few companies have been leading the development and innovation of AI and they hold onto vital information concerning the technology. The key regulatory challenge is how to prevent the inner workings of AI from becoming a black box to the public. The government should actively seek out what role it can play to achieve this goal.

It is also a time to review and reassess the general approach to regulating IT (information technology) giants and create an overall roadmap for the future.

--The Asahi Shimbun, April 6