Japan: New guidelines to encourage businesses to consider human rights when developing generative AI
"How does generative AI learn? Japan prepares disclosure rules", 12 August 2023
Japan will urge businesses to disclose the types of data used to train generative artificial intelligence programs, draft government guidelines reviewed Friday by Nikkei show.
The government hopes the guidelines, to be finalized later this year, will root out bias and other abuses that could be linked to AI...
Tokyo is hesitant to impose strict controls, given the speed at which generative AI is being developed. The guidance will not be legally binding, and it will carry no penalties for noncompliance.
Tokyo intends instead to offer norms for businesses to use when developing and using generative AI. To encourage adoption of the guidelines, Japan will establish a certification that companies are following best practices. Policymakers also envision requiring that businesses adhere to the guidelines when bidding for government contracts...
Businesses are to develop services that abide by the rule of law and consider human rights, with managers urged to establish internal risk management structures...
Knowing the nature of the original data would help AI service providers make better decisions about whether to adopt a program... Tokyo will explore the use of external audits that ensure businesses are maintaining transparency...