Starting Thursday, Naver's AI-based detection system will scan articles across all news sections — including politics and elections — and shut down commenting when abuse levels breach a set threshold.
The AI detection system Naver introduced in 2019 was the first in the domestic internet industry, continuously upgraded to flag negative expressions such as profanity, sexual, and violent language. The company said a further upgrade to the AI model is planned for later in April.
"Naver has been working to make the comments section a space for healthy communication, and we will continue to listen to diverse opinions to respond to rapidly evolving forms of malicious commentary," said Kim Su-hyang, a leader at Naver.
The move builds on a restriction Naver imposed last month, suspending comments beneath articles in its politics and election sections ahead of the June local elections. The company has also separately rolled out a "memorial comment" feature since February, allowing users to post condolences with a single click on articles covering disasters, accidents and deaths.
About 23 media organizations have adopted the memorial comment function to date, and articles featuring it have drawn about six times more comments relative to page views compared with other articles from the same outlets, Naver said.
Copyright ⓒ Aju Press All rights reserved.