Lawmakers Flag Minors' Safety on AI Chat Apps Zeta, Crack
At an October 12 National Assembly audit, lawmakers said Zeta and Crack generated self-harm and sexual exploitation content despite protection modes.
South Korean lawmakers raised concerns over youth protection on AI character chat services Zeta and Crack during a National Assembly audit on October 12, according to digitaltoday.co.kr.
Rep. Lee Ju-hee said Zeta's total monthly usage time is about 110 million hours, more than twice that of ChatGPT. Rep. Kim Seon-min said a test found that self-harm and suicide-related conversations were generated even with "Safe Mode" enabled on an account that had not been adult-verified. Rep. Hwang Jeong-a said sexual violence and sexual exploitation scenes involving minor characters were generated on Zeta without adult verification.
Kim Seon-min also noted that of 119 complaints related to AI character chat, 76 remain under review, and that a "BabyChat" report has been under review for 494 days.
Hwang Yong-seok of the Korea Internet Self-Regulation Policy Organization said self-regulation guidelines for AI character chat are being prepared, with the protected age provisionally set at 19.
Hwang Jeong-a said Scatter Lab has begun upgrading harmful content blocking features and plans to introduce youth usage time limits in the fourth quarter of this year.
Quick answers
What AI chat services were discussed at the National Assembly audit?
Lawmakers raised concerns about youth protection on AI character chat services including Zeta and Crack.
What did the audit reveal about protection modes?
Rep. Kim Seon-min said a test found that self-harm and suicide-related conversations were generated even with "Safe Mode" enabled on an account without adult verification.
What measures are planned to protect minors?
Scatter Lab has begun upgrading harmful content blocking features and plans to introduce youth usage time limits in the fourth quarter of this year, according to Rep. Hwang Jeong-a.