These days, 'Saturday Discussion with Jung Jun-hee' hosted by the Roh Moo-hyun Foundation showcases the most in-depth discussions and alternative solutions on major social issues and policies.
This week, we highly recommend the episode featuring Dr. Kim Hyun-jun and Im Mi-hyeon, co-director of the Solidarity for Participatory Democracy. Many people should watch and share it.
Here are a few points I'd like to emphasize:

1. Repeating social debates and problems in the legal process after the AI Basic Law

has been debated for over five years since the 21st National Assembly, with countless bills proposed by lawmakers and numerous public discussions. Ultimately, consensus was not reached within the 21st Assembly session, and it carried over to the 22nd Assembly.
- In the 22nd Assembly, the 'National Human Rights Commission of Korea' officially recommended supplementing the law due to insufficient human rights protection. This provided a strong impetus for the government to listen to civil society demands. As a result, continuous joint reviews and public hearings were held, allowing for deliberation and debate. It was widely anticipated that the law would be revised and enacted after another year.
- However, Yoon Suk-yeol unexpectedly triggered internal strife in early December, drastically changing the situation. With the national economy plummeting, all emergency measures became necessary, including a special investigation and impeachment proceedings. Consequently, an emergency meeting between the leaders of the two major parties was held, and the passed the National Assembly plenary session abruptly at the end of December.
- Under the pretext of economic recovery, most clauses related to human rights protection and essential regulations were deleted. Moreover, key provisions were largely delegated to the government through presidential decrees. The implementation date for these regulations was further postponed by another year.

1) Now, it's common knowledge that the core of South Korea's 'Miracle on the Han River' was the simultaneous success of democratization and industrialization. However, this crucial aspect has been completely overlooked in AI-related legislation.
Despite a change in government and rapid growth in stock prices and economic growth rates driven by large corporations, policy decisions are still heavily skewed towards the demands of some conglomerates. The misconception that "regulations hinder economic growth" persists, and AI legal policies are further accelerating this trend.
Regulations themselves are neither right nor wrong. Essential regulations can act as a filter for new industry growth and become a source of convergence and innovation. As our country has already witnessed significant progress in social cooperative economics, venture companies and startups have thrived by leveraging the 'redline' created by regulatory reforms in the past. Nevertheless, current AI legal policies in Korea continue to regress.

2) In a democratic and globalized society, this approach is far from efficient. It ultimately leads to delayed resolutions, increased social conflicts, and constitutional petitions, eroding public trust in government policies. As discussed in the debate, a shift in perspective and attitude is urgently needed.

2. The Need for Prompt Enactment of the National Defense AI Act

Despite these concerns, the enactment of the National Defense AI Act should be expedited.
- First, the exempts national defense AI from its application scope, creating a pressing need to establish specific regulations through separate legislation.
- Furthermore, with escalating geopolitical tensions and the growing threat of war, major powers surrounding South Korea are rapidly transitioning to AI-powered weapon systems. We must respond accordingly.
The acquisition of new weapons typically takes over 10 years under outdated procedures. However, like other countries, establishing a fast track or simplifying unnecessary procedures for essential AI system adoption seems necessary.
- This reform can also contribute to addressing internal strife. Implementing AI requires the adoption of data clouds, promoting transparency in information disclosure and a shift towards science-based governance.
The South Korean military and National Intelligence Service still lack a complete understanding of how many individuals supported the internal strife. Some speculate that over 50% of senior officials support Yoon Suk-yeol. The whereabouts of notebooks containing sensitive information and gatherings at Lotteteria or Anga remain unknown.
- Connected to this are ongoing astronomical defense industry corruption scandals. Many retired generals and high-ranking officials have lobbied for US military contractors like Lockheed Martin, influencing the military's dependence on foreign weapons and hindering national defense sovereignty. Concerns persist about the influence of powerful lobbies advocating for the adoption of American systems like Palantir or Anduril instead of developing sovereign AI capabilities for national defense.

3. Key Issues and Challenges in the Current National Defense AI Act

1) To gain public support and build consensus on the necessity of the National Defense AI Act, it is crucial to explicitly enshrine human control and democratic oversight in the legislation.
Frankly, the current bill submitted to the National Assembly merely mirrors the Republican Party under Trump or big tech cartels aligned with them. As pointed out in Saturday Discussion, the bill lacks fundamental requirements and fails to reflect research and current trends on relevant issues both domestically and internationally.

2) Concurrently, active and leading participation in international norms and multi-stakeholder dialogues is essential. President Lee Jae-myung's UN and ASEAN speeches and statements during his early term exceeded expectations by strongly advocating for democratic multilateral cooperation and mandatory human control. However, recent developments like the San Francisco Declaration, the Data Center Special Act, and the National Defense AI Act seem to contradict this direction.
Meanwhile, the European Union and major US cities with thriving AI economies, such as New York, California, and Seattle, are continuously strengthening their leadership by enacting sector-specific regulations. Moreover, they collaborate closely across the Atlantic, coordinating implementation timelines and regulatory content. It is unclear whether South Korea's AI diplomacy is actively engaged with these efforts.

A meeting of the UN Convention on Certain Conventional Weapons (CCW) Group of Governmental Experts on Lethal Autonomous Weapon Systems (LAWS) will be held for the first time in five years this October. This presents a historic opportunity to adopt a binding and authoritative treaty protocol banning or regulating 'AI autonomous weapons systems'.
However, South Korea has been identified by international NGOs as a key country obstructing the adoption of such a protocol over the past decade. Our responsibility for perpetuating this bleak future outlook is immense.
We urge the government and National Assembly to reflect on their actions and strengthen their commitment to responsible AI governance.

▶ Original Source: https://youtu.be/_jA4DYa_vcY?si=FlaONzRvDvRa_PsI