Naver's 'AI Cleanbot' Now Filters Excessive Sexual Expressions Too

[Asia Economy Reporter Kang Nahum] Naver's AI Clean Bot, a program that blocks malicious comments, filters out not only profanity but also excessive sexual expressions.

Naver announced on the 29th that the AI Clean Bot was upgraded to version 2.5 after seven months since June last year.

For this upgrade, the AI Clean Bot utilized about 160 million data entries, including previously learned data, over four months for training and enhancement. As a result, its accuracy improved by approximately 15% compared to before.

Notably, version 2.5 incorporates AI training to detect excessive sexual expressions. By blocking expressions that may cause sexual discomfort, the scope of malicious comment filtering has been expanded. From now on, Clean Bot will automatically block direct or indirect sexual acts or expressions related to specific body parts.

Along with this, Naver has taken measures to improve user accessibility for reporting in line with the enforcement of the so-called 'n-beonbang Prevention Act,' which prohibits the distribution of illegal filming materials. On the 15th, the 'illegal filming' report category was added to various areas including News, Cafes, Blogs, and KnowledgeIn, and help information for the reporting center is displayed during searches related to 'illegal filming.'

A Naver official stated, "Beyond declarative meaning, we will make our best efforts to practically block malicious comments and prevent secondary damage from illegal filming materials."

This content was produced with the assistance of AI translation services.