CI Leaks and AI Chatbot Overuse Put User-Protection Blind Spots Under Scrutiny at Parliamentary Audit of Media Commission

[Financial News] As personal data breaches continue to occur, calls are growing to improve the system so users can directly request renewal of Connecting Information (CI), which is difficult to change even after it has been leaked. After the need to introduce a right to request CI renewal was raised at a parliamentary audit, the Broadcasting, Media and Communications Commission said it would review related legal improvements and submit a proposal before the comprehensive audit. It also plans to pursue legal revisions to ensure transparency in YouTube’s recommendation algorithms and introduce safeguards to prevent teenagers from excessively using artificial intelligence (AI) chatbots.
At the National Assembly of South Korea’s Science, ICT, Broadcasting, and Communications Committee’s audit of the BMCC on the 12th, lawmakers focused on guaranteeing users’ rights after CI leaks, gaps in oversight of online platforms’ recommendation algorithms, and measures to protect teenagers using AI chatbots.■ CI is difficult to change even after a leak... Calls to introduce the right to request renewalCI is information used to identify the same user across multiple online services, and its leakage could lead to it being misused to link information. Rep. Kim Seon-min of the Rebuilding Korea Party noted that after the CIs of 1.29 million people were leaked from Lotte Card last year, 13.24 million cases were reported at TVING this year, and CI leaks had been confirmed at seven organizations in the past year alone.
Rep. Kim took issue with the fact that users find it difficult to change their CI or stop using it even after a leak. In one case, a victim asked six companies to stop processing their CI, but some companies reportedly refused or advised the person to close their account.
Rep. Kim called for an alternative process for users who do not wish to use CI and said users’ right to request CI renewal should be guaranteed separately from the validity-period system, which the BMCC is reviewing with validity periods of one to three years.
BMCC Chair Kim Jong-cheol said, “We are preparing legal improvements related to the issue of Connecting Information,” adding that the commission would consider measures including the right to request renewal. He also agreed to a request to submit a plan for improving the system before the comprehensive audit.■ Gaps in algorithm regulation... Protecting teenagers from AI chatbots also a key issueGaps in oversight of recommendation algorithms on online platforms such as YouTube were also highlighted. Rep. Kim Woo-young of the Democratic Party of Korea argued that recommendation algorithms could fuel the spread of misinformation and conspiracy theories by repeatedly exposing users only to information aligned with their preferences. He particularly noted that regulatory blind spots are emerging because it is unclear which government ministry is responsible for overseeing the algorithms.
Rep. Kim said that no department takes responsibility for the harms caused by algorithms, with departments saying, “That’s not within our remit,” and called for the relevant ministries to establish a coordinated response system.
Chair Kim said, “We have a duty to formulate and implement policies to address the harms caused by algorithms as part of protecting users and establishing the public’s media sovereignty.” He added, “We plan to work to protect users’ right to choose and ensure algorithmic transparency through revisions to the Telecommunications Business Act and the Information and Communications Network Act, among other measures.”
Concerns were also raised about harmful responses from AI chatbots and teenagers’ excessive use of them. Rep. Lee Ju-hee of the Democratic Party of Korea pointed to the risks of “companion AI,” in which people talk with AI characters as if they were friends or romantic partners.
According to materials presented by Rep. Lee, the AI character chatbot Zeta was used for about 110 million hours per month, more than twice the usage time of ChatGPT. Nearly nine out of 10 users are in their teens or twenties. A survey cited by Lee also found that 52% of teenagers who asked AI about self-harm, suicide, violence, or sexually explicit content received specific responses.
Rep. Lee called for measures informed by examples from countries including the United States and Canada, such as age verification and parental consent, limits on usage time and addictive design, and systems for responding to risky conversations. Chair Kim said, “Through consultations with the relevant ministries, including revisions to the Information and Communications Network Act or improvements to the AI Basic Act, we will do our utmost to ensure these measures are fully implemented.”
[email protected] Choi Hye-rim Reporter