The current legislative actions to regulate online hate speech in South Korea seem to have been only half completed

Executive Summary

• Problem statement: There is too much hate speech on Korean online platforms which exposed almost 40% of online users to hatred or cyberbullying
• Criteria:

1. Primary goal

- Fix biased attitudes to build a clear and safe online space where no hate speech exists

2. Secondary goals

- Develop a united understanding of online hate speech and re-define it in the Korean context

- Enable internet companies to organize and implement their own community guidelines

- Build an internet governance system to swiftly implement the hate speech regulation

• Recommendation:

- 1st: Hold a public discourse with the most important actors and publicize its result

- 2nd: Focus on damage recovery by building centers for remedy and counseling for victims

- 3rd: Repropose the legislation of the online hate speech regulation

- 4th: Strengthen internet etiquette education to reduce online hate speech generation

Background

There is too much hate speech on Korean online platforms which exposed almost 40% of online users to hatred or cyberbullying ⅰ). Domestic online platforms (e.g. Naver, Daum) have shut down their comment section on account of the significance of the problem. However, a number of online extremists, such as ‘Cyber Wreckers’ ⅰⅰ), have sprung up on foreign internet platforms (e.g. YouTube, Facebook, Twitter) and endlessly create groundless slander toward a certain individual or a group. Even worse, traditional media is uncritically taking the dictation of what the wreckers said without proper fact-checking. It has led not only to severe cyberbullying and even the suicide of celebrities (e.g. Sul-li, BJ Jammi, Volleyball player Kim In-hyuk), but also to indiscriminate hate crimes that have recently occurred in Korea ⅲ).

The root cause of the online hate speech problem can be explained through the Pyramid of Hate ⅳ). It clearly shows that biased attitudes (e.g. stereotypes, fear of differences) lead to actual discrimination and violence both online and offline. Furthermore, malicious online content, comments or real hate crimes are spread very fast through the internet by media and SNS, reinforcing the biased attitudes against a certain person or group again which then bring about a larger scale of aggression. This ‘self-reinforcing’ vicious cycle ⅴ) indicates the wickedness of the problem.

Stakeholders

In regard to the online hate speech problem, there are twelve stakeholder groups under three categories. Using the political analysis worksheet suggested by David L. Weimer & Aidan R. Vining (2015), I specified each stakeholder group’s motivations, beliefs and resources in an appendix at the end of this memo. This might help you to understand potential clashes or common areas between stakeholder groups.

Criteria for policy success

Legitimacy: It is worth noticing that the majority of prior legislation (including your effort) to regulate online hate speech ended up in failure because of the lack of legitimacy & support ⅵ). Despite the robustness of the public value it pursued, it was suspected to be likely to infringe freedom of speech. Even though the UN clearly contradicted this argument ⅶ), it seems evident that the authorities have been failing to convey this message to the general public. Therefore, it is essential to reach a social consensus as a means of laying the groundwork for successful legislation.

Operational Capacity: Platform companies might not have sufficient motivation or technical capabilities to strictly implement their online community guidelines. In the case of Youtube, advanced machine learning technology is utilized to flag and remove online violations. But this is just an instrument to sift out pebbles from sand seeing as 500 hours of video are uploaded every minute. Moreover, the last decisions for content removal are yet subject to human reviewers who are costly to hire and maintain as well as might be biased.

Urgency: The urgency of the problem needs to be considered significantly. Given the nature of the internet, online hate speech is circulated at a much faster rate and extremely manifests more than offline hate speech and hate speakers are able to hide behind anonymity. This is why it is impossible to perfectly prevent its victims from suffering or even dying while the authorities implement the best mix of solutions.

Mental Models: The most idealistic situation would be an online space where no hate speech is generated at all. If there is no hate speech circulating on the internet, there is no reason to regulate it. This is why fixing the mental models (e.g. biased attitudes in the lowest of the Pyramid of Hate) is the most fundamental and effective solution to end this never-ending vicious cycle.

< Criteria for successful policy implementation >

• Primary goal (Mental Models)

• Fix biased attitudes to build a clear and safe online space where no hate speech exists
• Secondary goals (Legitimacy, Operational Capacity, Urgency)

• Develop a united understanding of online hate speech and its harmful effects and re-define it in the Korean context among relevant stakeholders (Legitimacy)

• Enable internet companies to organize and implement their own guidelines with the support of a legally binding common standard and technical assistance (Operational Capacity)

• Build an internet governance system (between public sector, private sector, civil society) to swiftly implement the regulation when it is socially acceptable (Urgency)

Policy options

1. Hold a public discourse with the most important actors and promote its result extensively

To legitimize the online hate speech regulation, it is necessary to reduce a perception gap that clearly exists between the general public and expert groups ⅷ). In addition, an event organizer must include internet service providers, content creators, online users and if possible, other political parties who used to be marginalized from this type of public hearing.

2. Repropose the legislation of ‘Online Violence Prevention Law’ and/or ‘Anti-discrimination Law’

Or, you can reattempt to obtain the legislation of the laws you suggested. This is because the tenure of the current members of the National Assembly is very limited and ends on May 30 2023. Realistically, it is possible to encounter identical problems that they had in the previous failures. But, in the meantime, if there are pivotal turning points triggered by civil society, the proposals with a few revisions may be able to be passed in the next session.

3. Strengthen internet etiquette education to reduce online hate speech generation

Considering the Pyramid of Hate, changing the mental model (biased attitudes) is the most fundamental solution to foster an online environment where no hate speech is produced. Although it has a long way to go before seeing its effectiveness, offering proper education or holding campaigns can be the surest way in the long term to encourage every online user not to produce any hate speech.

4. Focus on damage recovery by building centers for remedy and counseling for victims

You can put an emphasis on after the fact responses. For example, you might be able to co-organize a facility with government administrators and NGOs to support the victims of online violence by referring to Korea Cyber Sexual Violence Response Center ⅸ) as a role model.

5. Maintain the status quo by trusting the current internet ecosystem’s self-purification ability

Relying on the current community guidelines and user report systems, the authorities might not have to do anything. According to Kim Un Kyung, a media communication professor of the HUFS, the guidelines made by Korean internet companies are largely well-organized with specific instructions ⅹ). Considering the side effects of the legislation (e.g. political backlashes ⅺ), potential abuses ⅻ)), doing nothing can be a solution with the trust on the ‘self-correcting’ effect of the private sector and civil society.

Recommendation

I highly recommend you to carry out the 1st and 4th options initially, and then concentrate on the 2nd and 3rd options subsequently. The previous legislative failures were mostly attributed to the lack of social consensus on defining what online hate speech is and why it matters in Korea. Thus, it is very important for every member of society to comprehend the problem in depth. It won’t be too late to come up with more efficient regulatory methods (e.g. imposition of legal duties) after that. In addition, victim support must be executed without hesitation given the urgency of the problem.

Nevertheless, we shouldn’t downplay the role of prevention measures. Even though time is running out, our final destination should be to make a clear and safe online space with no harm. This would be realized by switching the mental model (alleviating biased attitudes) through the means of providing education and holding campaigns.

Thank you for taking your time to read this policy memo. I appreciate your hard work on this matter and hope it helps you to be oriented in a more precise direction.

Endnote

ⅰ) KCC (2023). 2022 Cyber Violence Review Report. Korea Communications Commission.

ⅰⅰ) According to Lee Shin-haeng & Hyun A-yeon (2023), they are an emerging type of personalized media content producers who post videos about celebrity scandals or incidents.

ⅲ) In the case of a serial murderer who committed a crime in Seo-hyun Station, he foreshadowed a crime in online community called, ‘DC Inside’, where most online hate speech in Korea is generated (An Jung-hoon, 2023).

ⅳ) According to the Pyramid of Hate, biased attitudes such as stereotypes or fear of differences lead to the acts of bias, discrimination, bias motivated violence and even genocide (Anti-Defamation League, 2021)

ⅴ) Hatred basically has a trait of a positive feedback loop, which refers to a self-perpetuating pattern of behavior where the final result is intensified by the initial act. (Akhilesh Ganti, 2022 & Yudkin, J.S., Messiah, S.E., 2019)

ⅵ) According to the Strategic Triangle presented by Mark H. Moore, to implement a policy successfully, policy makers are required to meet three requirements: 1. Legitimacy & Support  2. Operational Capacity  3. Public Value (Mark H. Moore, 2017)

ⅶ) The UN Secretary General clearly concluded that addressing hate speech does not mean confining or prohibiting freedom of speech. It means preventing hate speech from escalating into something more dangerous, particularly incitement to discrimination, hostility and violence, which is banned under international law. (United Nations, n.d.)

ⅷ) The people (e.g. online users) tends to view online hate speech as a threat to individuals (e.g. celebrities, sport stars) whereas expert groups (e.g. Members of the National Assembly, KNHRC) tends to view it as a potential attack towards social minorities (e.g. women, the disabled, the elderly) (The Blue House National Petition, 2022 & Hate speech report, 2019).

ⅸ) Korea Cyber Sexual Violence Response Center is an organisation that supports the victims of online sexual violence through counselling, video deletion and legal assistance (KCSVRC, n.d.).

ⅹ) Neglected Hatred: What can we do to solve online violence? (2022, Feb 17). The Korean National Assembly Member Jang Hye-yeong’s urgent discussion session materials. Justice Party of South Korea.

ⅺ) After a few members of the Korean National Assembly proposed the Comprehensive Anti-discrimination Law, there were a number of assemblies and demonstrations arguing that the regulation of hate speech is likely to damage the value of freedom of speech (Richard Kim, 2023).

ⅻ) One of the side effects of the legislation of online hate speech regulation would be a possibility that a man of power abuses this law to suppress his opposite parties and gain more political power.