Discussion on AI Regulation
The government of Japan embarked on legislating a law regulating development of artificial intelligence (AI). While the government has been reluctant to introduce regulation on AI, expecting momentum for economic growth, the developed countries are promoting the regulation, as if leaving Japan behind. Japan’s measure for controlling Al, however, is a mixture of measures for protecting individuals and supporting technological development.
AI Strategy Conference, a discussion body of the Cabinet Office, raised some points for discussing how to deal with AI. According to “Thoughts on System for Dealing with AI,” a document submitted to the conference, the government recognizes that the opportunity and availability of AI have been expanding with development of generative AI. AI has been a tool for economic growth.
However, the risks that AI embraces became unignorable in modern society. Malfunction of AI in medical devices or vehicle for transportation directly affects human life. Wrong information in surveillance brought by AI may violate human rights in such case as disseminating fake pictures. Apparently, AI can be used for military weapons, terrorism or cyber attacks, endangering democracy.
“Can voluntary discipline be enough for reducing the risks?” “Don’t we need appropriate regulation for healthy competition?” Those are the questions the conference raised. The conference is seeking to maximize benefit of AI for creation of innovation, solution of social problems, keeping international competitiveness and improving people’s quality of life. “Appropriate action on the risks through clear discipline is needed as a ‘guardrail’ for development, supply and utility of AI,” said the document.
To the high risk developers of AI, the conference considers regulation system aiming at improvement of safety and security of the people by enhancing safety and credibility of AI, from the perspectives of protecting life, liberty, property and basic human rights of the people. To the low risk developers, including small-size entrepreneurs or researchers, the regulation will take a soft approach.
Some lawmakers in the Liberal Democratic Party made in February a draft of “Responsible AI Promotion Basic Act,” which required appropriate governance for minimizing the risks and maximizing the benefit. Ministry of Economy, Trade and Industry issued in April AI Guidelines for Business to urge AI businesses to recognize the risks and to take necessary measures.
European Council approved AI Act, the flagship legislation following a risk-based approach, which meant the higher the risk to cause harm to society, the stricter the rules. US President Joe Biden issued an executive order last October, which was designed to protect American citizens from the risks of AI. AI Strategy Conference of Japan argues that the rule in Japan needs to be relevant to international regulation system. Japan takes action when it realizes international pressure.
Comments
Post a Comment