Naver Unveils AI CleanBot 3.0 to Prevent Content Devaluing Life and Causing Secondary Harm

Naver deployed an upgrade to its 'AI CleanBot 3.0' on the 29th, designed to block malicious comments. This upgrade specifically focused on enhancing the system's ability to detect and block expressions that trivialize the value of life or cause further distress to victims. Unlike previous methods that solely analyzed comment content, the system's performance has been improved to more accurately identify malicious intent by analyzing both the article title and body content together.

Since its industry-first introduction in 2019, Naver has consistently enhanced the detection range and accuracy of AI CleanBot to respond to new slang and hate speech. In 2020, it added sentence context judgment capabilities and strengthened its function to block circumventing malicious comments using symbols or characters. In 2023, it continued upgrades by reflecting the KISO (Korea Internet & Self-Regulation Organization) self-regulation guidelines for hate speech.

Kim Soo-hyang, a Naver leader, stated, "We are continuously strengthening CleanBot's performance to detect not only profanity and vulgar language but also newly emerging expressions of hate, demeaning speech, and discrimination." He added, "We plan to further refine CleanBot's performance by focusing on blocking content that promotes disregard for life, mocks victims and their families, and spreads hate, while also listening to various opinions."