• NTT and Yomiuri Shimbun propose legal and technological measures to control the impact of generative AI (GenAI)
• Unrestrained development risks undermining social stability
• The partners suggest those companies interacting directly with GenAI users need to take responsibility to mitigate risks

As generative artificial intelligence (GenAI) technologies rapidly advance, concerns are growing about the potential risks they pose to individuals and society as a whole. In a joint proposal released earlier this month, giant Japanese telco NTT and media company Yomiuri Shimbun Holdings outlined the challenges and proposed solutions for responsibly shaping the development of GenAI.

The proposal, for which studies were initiated during the second half of 2023, highlights the advantages of GenAI, such as its user-friendly interfaces and potential to improve productivity. However, it also warns that the current inability to fully control the technology could lead to serious consequences, including the erosion of trust in society and even the collapse of democracy and social order in worst-case scenarios.

Central to the issue is the “out-of-control relationship between AI and the attention economy,” which the companies argue has already damaged individual autonomy and dignity. The proposal states, “If generative AI develops further and is left unchecked like it is currently, there is no denying that the distribution of malicious information could drive out good things and cause social unrest.”

To address these risks, the companies propose a combination of legal restraints and technological solutions. They call for strict restrictions by law on the use of GenAI in sensitive areas, such as elections and security, as well as careful measures in education. The protection of intellectual property rights, particularly copyrights, is also identified as a key area for adaptation.

The proposal emphasises the need for effective governance, including legislation and co-regulatory approaches that involve stakeholders. It suggests that businesses interacting directly with users in the value chain should play a role in reducing and absorbing risks associated with GenAI.

“Generative AI is a technology that cannot be fully controlled by humanity. However, it is set to enter an innovation phase,” the proposal states. “Measures to ensure a healthy space for discussion, which constitutes the basis of human and social security (democratic order), must be taken immediately.”

NTT, which is developing its own large language model (LLM) called Tsuzumi, and Yomiuri Shimbun have pledged to continue studying the issue and making recommendations, with support from the Cyber Civilization Research Center at Keio University. As the proposal concludes, “The most important thing is to protect the dignity and liberty of individuals in order to achieve individual autonomy.”

- Joana Bagano, Contributing Editor, TelecomTV