Skip to main content

Japan to apply AI techniques to monitoring the Net

Very interesting to see that, in the aftermath of the horrific stabbing spree by Tomohiro Kato in Tokyo last weekend, the Japanese government has outlined plans to use artificial intelligence to monitor what is happening on t'Internet.

In case you were wondering, the 25-year-old, was arrested last Sunday after ramming a truck into pedestrians in a crowded part of Tokyo and slashing people at random with a dagger, killing seven people and injuring 10 more.

Kato left hundreds of messages on the Net saying he felt lonely, ugly and angry. And as he drove to Tokyo on Sunday morning, he sent more postings via his mobile spelling out his plans for a massacre.

"I'll crash my vehicle into people and if the vehicle becomes useless, I'll get out a knife. Goodbye everyone!" said one posting hours before the attacks started, said one Japanese newswire.

The Ministry of Information in Japan has said it already has software that can detect certain words when someone posts them online, so extending the software is not a major job.

The problem, however, says the Ministry, is that searching for the keyword `murder' generates a lot of information to screen in real time, so AI may be the answer.

The new software will seek to "catch when somebody's messages escalate to an alarming level in certain contexts."

The fact that the Japanese government is a lot more open about its security that our secret squirrels here in the West makes me wonder of the NSA and/or MI6 hasn't developed something similar here.

I suppose we'll never get to find out...