TOKYO — As reports of child abuse continue to increase, the Mie prefectural government plans to use artificial intelligence from fiscal 2019 to help decide whether temporary custody is necessary to protect children.
The project is supported by the National Institute of Advanced Industrial Science and Technology (AIST).
AI-equipped devices will be provided to some of the prefecture’s local field offices.
The use of AI, based on accumulated data, is expected to accelerate the decision-making process and reduce the staff workload at the offices.
Studying the potential use of AI in measures against child abuse was included in the central government’s emergency action plan following the abuse-related death of 5-year-old Yua Funato in Meguro Ward, Tokyo, in March 2018.
If introduced, Mie Prefecture will be a pioneer in using AI to handle child abuse cases in the field.
In fiscal 2013, the prefecture developed a “risk assessment sheet” that lists points to confirm when deciding the necessity of temporary custody.
The AIST analyzed data from about 5,000 sheets to develop the AI-equipped device.
When an official enters data into the device in the same categories as those on the risk assessment sheets, AI will analyze the case’s characteristics and provide information, such as the predicted likelihood that abuse will occur again.
A camera on the device can also take photos of marks on a child’s body.
As the system continues to accumulate data, it will become possible for AI to determine whether the mark in a photo was caused by abuse.
The AIST will provide about 20 such devices for a model project in the first year.
“With the help of science and AI, we want to deal with circumstances that would be difficult to anticipate using our experience only,” a prefectural official said.