In this Sept. 30, 2015, photo, South Korean pop star and actress Sulli poses during the K-Beauty Close-Up event in Seoul, South Korea. Jang Se-young/Newsis via AP
In South Korea, malicious online comments that defame others were once regarded as a matter of personal conduct and thus were not taken seriously and regulated.
However, in 2019, the suicide of South Korean singer and actress Choi Jin-ri, better known as Sulli, was widely discussed in connection with the cyberbullying she had endured for years. (Park & Kim, 2021, p. 1)
Following her death, South Korean society began to discuss the governance of malicious comments and renewed calls for real-name requirements, attempting to curb cyberbullying. (Park & Kim, 2021, p. 3)
However, six years later, in February 2025, South Korean actress Kim Sae-ron also took her own life. In the approximately 1,000 days following her drunk driving incident, South Korea’s mainstream media published around 2,000 related reports, placing her under sustained public scrutiny. (Kim Tong-Hyung, 2025)
South Korean actor Kim Sae-ron arrives at the Seoul Central District Court in Seou,South Korea, Wednesday,April5,2023.(Yonhap via AP)
Even though South Korea has already imposed restrictions on malicious comments and individual behaviors, such an incident still occurred, indicating that the problem is not merely about malicious comments.
The deeper problem is the harm constituted by the platform, media, anonymity mechanism and the economy of attention. To address online harm, efforts should not merely focus on restricting content and behavior, but rather shift towards the entire system that make harm scalable.
That Was Never Just “Mean Comments”
If online violence is simply reduced to malicious comments, the way to deal with it would naturally become simple: just delete the posts, mute or ban the accounts, and add a sentence in the community asking for civilized speech.
However, the problem is not just that. Online violence is never constituted by a single comment, but accumulates over time.
Sinpeng et al. (2021) frame hate speech not as speech that “merely offends someone, or hurts their feelings,” but as speech that “can harm immediately and over time” (p. 6).
A single comment may seem harmless, but when tens of thousands of such comments are all piled up on one person and last for years, it becomes a continuous form of harm and violence.
This phenomenon is very prominent in South Korea.
Park & Kim (2021) once pointed out: “online misogyny is so prevalent that more than 80% of respondents… reported having been exposed to misogynistic language” (p.2)
This indicates that when South Korean women go online, being attacked is no longer a surprise to them; it’s almost like a daily occurrence.
Sinpeng et al. (2021) also indicate that the true harm of hate speech lies in the deprivation of powers it causes (p. 12).
The attacked will gradually lose the ability and willingness to speak out on an equal footing, and thus lose the status of being treated as a human being. Therefore, the more pressing issue that needs to be addressed in the governance of cyber violence is the problem of power structure. It is necessary to further explore how this deprivation of power is magnified by platforms and the like.
What Changed After Sulli’s Suicide
After Sulli’s death, Naver did not simply return to a full real-name regime. Instead, it adopted an intermediate model of accountability.
Park & Kim (2021) mentioned in their article that “decisions by the country’s leading news providers (Naver and Kakao) to eliminate the comments section that accompanies ‘entertainment’ news were made only after the suicide of Sulli” (p. 2).
Kang et al. (2022) documented three significant and crucial interventions by Naver in the spring of 2020:
Naver’s three major interventions in spring 2020:
– February 19: Naver announced self-regulatory measures.
– March 6: Naver removed the entertainment comment section entirely.
– March 19: Naver revealed all users’ comment history while keeping usernames pseudonymous.
(p. 468, table 1)
From this, it can be seen that Naver did not choose to return to the real-name system but instead adopted an intermediate state.
Users remain anonymous to the outside world, but all the comments they have made in the past can be traced.
This intervention has indeed worked. Kang et al. (2022) argue that “while user interaction on Naver has decreased, the percentage of negative comments has also fallen, showing that Naver’s regulations have had favorable effects” (p. 465).
After Naver made reforms to the comment sections of its entertainment sections and made users’ comment histories public, the situation of online harmful comments improved. (Kang et al., 2022. p. 472)
Among these, the step of making comment histories public was particularly crucial. The proportion of negative comments decreased after this, and the proportion of positive comments also increased. (Kang et al., 2022. p. 473)
The platform is not incapable of reducing violence on its own, but it needs to be pushed to do so.
However, Kang et al. (2022) also pointed out that “The removal of comments on Naver eliminated a channel for cyberbullying from Naver’s entertainment section, but not from Korean society as a whole.” (p. 474)
The suicide case of Kim Sae-ron in 2025 is a powerful proof of this.
Even if comments are improved, harm simply migrates to other spaces.
Harm Did Not End But Moved
Although closing the comment sections of entertainment news does indeed cut off a concentrated channel for malicious comments, as Kang et al. (2022) concede, the effect is still insufficient. (p.474)
The suicide of Kim Sae-ron in 2025 is proof of this.
Associated Press (2025) reported that from Kim Sae-ron’s 2022 drunk driving incident until her death, about 1,000 days later, South Korea’s mainstream news media published approximately 2,000 reports related to her.
This continuous media scrutiny is not a problem of a single comment section, as it occurred after comment sections had been managed and closed.
Kang et al. (2022) pointed out the migration between platforms: “News reports emerged of cyberbullies simply joining different social media platforms such as Instagram” (p. 474).
Park & Kim (2021) also support this point in their discussion and analysis of Twitter: “users also condemned uncivil news coverage prevalent in the Korean media and unethical journalistic practices” (p. 6). Users are not only blaming trolls but also pointing out that immoral entertainment news reporting and journalistic practices should also be criticized.
Cyber harm and violence are not just malicious comments but a vicious cycle between these, media coverage, and the amplification by social platforms.
Cover et al. (2024) found in their study of bullying and harassment policies on mainstream platforms like Facebook and Instagram that these platforms explicitly differentiate between public figures and ordinary users at the rule level (p. 2156). They discovered: “public interest concepts are actively co-opted by platforms to excuse differential policies that increase revenue” (p. 2160).
Each platform is actively formulating rules, which makes it easier for public figures to suffer online harm. What Sulli and Kim Sae-ron experienced was not just ordinary malicious comments, but rather systematic violence that was encouraged by the platforms’ policies that treat public figures as legitimate targets of attack.
Even if comment sections are closed, as long as platforms continue to use such rules and media continue to view such attacks as traffic and profitable content, the harm will only shift, not disappear.
Real-Names Were Not the Answer
Since anonymity emboldens people to do evil, let everyone be held accountable for what they say.
Therefore, people chose a simple solution, real-name registration.
Park & Kim (2021) observed that in the Twitter discussions after Sulli’s death, a large number of users did indeed call for legislation to require real-name verification (p. 6).
However, even so, South Korea’s attempt still ended in failure.
Even though the South Korean government implemented the Network Act in 2009, which required all users to verify their real identities before posting comments, this system was still abolished in 2012. (Kang et al., 2022. p.468, Table 1).
As Kang et al. (2022) points out, “the way the system handled real name information was flawed and that it did not impede hate comments” (p. 473).
This system overly restricts people’s freedom of expression and has apparently driven users to foreign platforms (p. 473).
Naver’s pseudo-anonymity design, launched in 2020, mitigates the excessive restrictions of strong anonymity while ensuring that every comment can be traced back to its source. Kang et al. (2022) note that Naver’s approach “learned from the shortcomings of the Network Act. Personal information is not exposed to the public” (p. 473).
The key to pseudo-anonymity lies in separating identity exposure from behavioral accountability, but the Network Act confuses the two.
Naver maintains the anonymity of speakers, but all past comments can be viewed through their accounts.
This indeed ensures that even if people protect their virtual identities online, they still need to be responsible for their remarks.
This also indicates that platforms can reduce harm by redesigning visibility, reputation, and accountability mechanisms without exposing everyone’s real offline identities.
Therefore, anonymity and lack of restraint are not the same thing.
Sinpeng et al. (2021) mention that the governance of hate speech requires “local knowledge to identify, and consultation with target groups to understand the degree of harm they experience” (p. 1), “Make transparent the types and weight of evidence needed to take action on hate figures and groups, to assist law enforcement agencies and civil society groups in collating this information” (p. 2), and “hate speech is very context-dependent, and intimate local knowledge is required to understand and address it fully” (p. 39).
Platforms should be held responsible for the environments they create and amplify that cause harm, which is the lesson brought by Sulli and Kim Sae-ron together.
What Better Platform Governance Should Look Like
The cases of Sulli and Kim Sae-ron demonstrate that addressing online harm cannot merely rely on managing malicious comments or closing comment sections.
Sinpeng et al. (2021) mentioned that hate speech is “structurally derived when the speaker enacts oppressive permissibility facts… speech that perpetuates systemic discrimination against the group to whom the target of hate speech is perceived by the speaker to belong.” (p. 11).
Within this framework, online harm has always been a systemic phenomenon rather than solely caused by malicious comments.
Even if managing comment sections can alleviate some harm, if platforms continue to amplify such harm and the media continue to treat it as profitable material, the cycle of harm will persist. Governance should view recurring harassment as a systemic risk.
Sinpeng et al. (2021) pointed out, “The language and context dependent nature of hate speech is not effectively captured by Facebook’s classifiers or its global Community Standards and editorial policy. It requires local knowledge to identify, and consultation with target groups to understand the degree of harm they experience.” (p.1).
Platforms need to identify such patterns, determine if there are targeted long-term attacks, and whether there is mutual amplification of hate speech among platforms. Public figures are not inherently fair game for attacks. Platforms and media should also take responsibility in this harmful environment, not just the haters.
Conclusion
Sulli committed suicide in 2019, while Kim Sae-ron did so in 2025. Despite the fact that six years have passed and South Korea has introduced various reform plans, the core system that enables the harm to be scaled up has never been touched.
Online harms cannot be reduced to individual malicious comments. The targeted and repeated attacks on an individual for over a thousand days and nights could not have occurred without the amplification effect of media coverage and the influence of platform amplification.
To control the harm of the internet, it is not enough to merely silence the public. The management of platforms and media as well as the protection of victims all need to be regulated.
The deaths of Sulli and Kim Sae-ron have highlighted the severity of internet harm and also reminded people that this is a governance issue that must be taken seriously.
Reference page
Cover, R., Henry, N., Huynh, T. B., Gleave, J., Grechyn, V., & Greenfield, S. (2024). Platform policy and online abuse: Understanding differential protections for public figures. Convergence: The International Journal of Research into New Media Technologies, 30(6), 2152-2167.
Jang, S.-y. (2015, September 30). [Photograph of Sulli at the K-Beauty Close-Up event in Seoul]. Newsis via Associated Press. https://apnews.com/article/south-korea-actor-kim-sae-ron-suicide-media-coverage-youtube-3924a3a118327d54cf703b1fc0e1e5e8
Kang, N. G., Kuo, T., & Grossklags, J. (2022). Closing Pandora’s Box on Naver: Toward Ending Cyber Harassment. Proceedings of the International AAAI Conference on Web and Social Media, 16(1), 465-476. https://doi.org/10.1609/icwsm.v16i1.19307
Kim, T.-H. (2025, February 21). Kim Sae-ron’s death underscores the huge pressure on South Korean celebrities. AP News. https://apnews.com/article/south-korea-actor-kim-sae-ron-suicide-media-coverage-youtube-3924a3a118327d54cf703b1fc0e1e5e8
Park, S., & Kim, J. (2021). Tweeting about abusive comments and misogyny in South Korea following the suicide of Sulli, a female K-pop star: Social and semantic network analyses. Profesional de la Información, 30(5), Article e300505. https://doi.org/10.3145/epi.2021.sep.05
Sinpeng, A., Martin, F., Gelber, K., & Shields, K. (2021, July 5). Facebook: Regulating hate speech in the Asia Pacific. Final Report to Facebook under the auspices of its Content Policy Research on Social Media Platforms Award. Dept of Media and Communication, University of Sydney and School of Political Science and International Studies, University of Queensland.
Yonhap News Agency. (2023, April 5). [Photograph of Kim Sae-ron arriving at the Seoul Central District Court in Seoul]. Associated Press. https://apnews.com/article/south-korea-actor-kim-sae-ron-suicide-media-coverage-youtube-3924a3a118327d54cf703b1fc0e1e5e8
Be the first to comment