Police Strengthen Public-Private Coordination to Prevent Lone-Offender Terror Attacks, Expanding Partnerships to Stop Crime

The Yomiuri Shimbun

6:00 JST, October 28, 2025

Japanese authorities are expanding partnerships between police and private businesses to prevent terror attacks by lone offenders — individuals who radicalize in isolation without belonging to any organization.

In the July 2022 fatal shooting of former Prime Minister Shinzo Abe, suspect Tetsuya Yamagami, 45, now indicted on murder and other charges, had built his homemade gun at home without anyone noticing. How can society detect early warning signs — in local communities or online — and stop such acts before they happen? Government and private entities continue to search for answers together.

Help from real estate industry

“If you notice anything suspicious — such as the smell of gunpowder or metallic noises — please contact us immediately.”

On Sept. 25, Kaoru Sakuma, the head of security at the Metropolitan Police Department’s Takenotsuka Station in Tokyo’s Adachi Ward, visited a local real estate company to ask for cooperation.

The MPD in May signed an agreement with three real estate associations to share information about suspicious activity in managed properties. Since then, police officers from Takenotsuka Station have visited around 20 real estate firms in the ward.

Their efforts are driven by the case of Yamagami, who allegedly made firearms and explosives at his home and in a rented garage. A neighbor in the same apartment complex recalled hearing “sawing sounds,” but no report was ever made to the police.

Authorities have now added five chemical items including sulfur — the ingredient used for gunpowder by Yamagami — to their monitoring list, requesting businesses to report unusual purchases.

Social media surveillance

Police are also strengthening ties with social media and internet-related companies. Ahead of the Upper House election in July, the National Police Agency asked the Japan branch of X Corp., formerly Twitter, for cooperation, providing a list of coded terms such as “56-su” (pronounced “korosu,” which means “to kill”) to help identify posts threatening political candidates and other high-profile individuals.

X responded to police requests for user information within a few hours. One man who wrote, “If [Kishida] comes [to campaign], he might not make it out alive,” was quickly identified and warned, in a reference to former Prime Minister Fumio Kishida. During the one-month period leading up to election day, the agency found 889 dangerous posts on X and other social media platforms.

An X spokesperson said the company “will respond [to police’s request] appropriately while respecting freedom of speech and expression.”

Online marketplace Mercari, Inc. and LY Corp., the operator of internet portal Yahoo! Japan and messaging app Line, have also received police input and banned from July onward the sale of empty shell casings that could be used for homemade bullets and of Chinese-made toy guns capable of firing live rounds.

Gathering warning signs

In April last year, police nationwide established a system to consolidate pre-incident information obtained during investigations or patrols by criminal, community-safety and local divisions into their security departments. In April this year, the Metropolitan Police Department created a new section in the Public Security Division to serve as a command center for lone-offender countermeasures. The National Police Agency is also studying the use of artificial intelligence to detect dangerous social media posts.

“Maintaining a balance between public safety and the protection of freedom and human rights is essential,” said Mitsuru Fukuda, a professor of crisis management at Nihon University. “The key challenge will be how to create an environment in which ordinary citizens feel comfortable reporting to or consulting authorities when they sense potential danger.”