NTT frets over the potential malevolence of AI

The announcement frames the matter by starting with this statement: “The accuracy of results from generative AI cannot be fully guaranteed at present, and unlimited use of generative AI could cause various problems for humanity and society. Humanity must take measures to balance control and use of generative AI from both technological and institutional perspectives.”

Their main concerns are “the out-of-control relationship between AI and the attention economy,” and the creation of “legal restraints to ensure discussion spaces to protect liberty and dignity.” It’s hard not to interpret those as a call for greater censorship of social media, which follows similar trends in the US and Europe in what will be a pivotal year of elections.

In its ‘outlook for the future’ the announcement says “Measures to ensure a healthy space for discussions must be taken immediately. Strong restrictions are especially needed for elections and the security field.” Japan’s laws state its next general election must take place by the end of October 2025.

However, another bullet point states “In this process, the most important thing is to protect the dignity and liberty of individuals in order to achieve individual autonomy.” It will be interesting to see how they square that stated aim with the call for ‘strong restrictions’.

The full list of concerns and proposed remedies can be accessed here. One paragraph in particular jumps out due to its portentous and apocalyptic tone. “If generative AI is allowed to go unchecked, trust in society as a whole may be damaged as people grow distrustful of one another and incentives are lost for guaranteeing authenticity and trustworthiness. There is a concern that, in the worst-case scenario, democracy and social order could collapse, resulting in wars.”

Last week South Korea launched an AI strategy council, shortly after the UK and US announced their own AI safety initiative. Meanwhile the EU reckons it’s setting the global standard for this sort of thing. The underlying driver for a lot of this seems to be concern that gen AI will empower bad actors to negatively influence public opinion prior to general elections. While that seems to be a legitimate concern, there’s also a very real danger that prophylactic measures will also suppress genuine and sincere speech online.