New Digital Order: A blueprint for mutual prosperity through AI governance in Korea

This post is the foreword written by Willy Cho, Area Vice President, Microsoft Korea for Microsoft’s report New Digital Order: A blueprint for mutual prosperity through AI governance in Korea. Read the full report in English and Korean.

At the forefront of a technological revolution, Korea’s journey with artificial intelligence (AI) epitomizes a harmonious blend of innovation and responsibility. As we embark on a new chapter in human history, our approach to AI is not merely about technological advancements but about sculpting a future that resonates with our societal ethos and global ambitions.

South Korea is on a fast track to achieve AI Everywhere, with instances of AI innovation manifesting across all facets of society. Key industries in Korea are at a transformative inflection point, rapidly embracing AI technologies, and Microsoft has been an indispensable partner in this journey.

Example 1) An exemplary case of this collaboration is in the field of senior healthcare, where we are addressing Korea’s challenge of an aging population. Together with Microsoft, Professor Howard Lee from Seoul National University’s Center for Convergence Approaches in Drug Development is conducting research on algorithms aimed at enhancing drug repurposing, which involves discovering new therapeutic uses for drugs, as well as AI-driven improvements to clinical trial design. The outcomes of this research include valuable insights into maximizing cost-effectiveness and improving the accuracy of clinical trials.

Example 2) Microsoft is also supporting research on deep learning for structure-based drug design, led by Professor Choi Sun at Ewha Woman’s University’s Global AI Drug Discovery Research Center. This research has succeeded in rapidly screening molecules with targeted properties and compounds with high binding affinity by analyzing protein and compound data with AI. Microsoft’s domestic AI research collaborations are contributing to significantly reducing the time and cost of drug development in healthcare and improving drugs’ medical efficacy, thereby easing the financial burden of managing geriatric diseases on the national healthcare system.

Example 3) More recently in March 2024, the potential of AI in educational settings has been explored through the lens of a project at the Korea Advanced Institute of Science and Technology (KAIST). Professor Alice Oh and her team embarked on a journey to harness the educational benefits of AI while mitigating the risk of it being used for shortcuts in academic work. Their solution came in the form of a chatbot developed with support from Microsoft Research’s Advancing Foundation Models Research (AFMR) initiative, launched in April 2023. This initiative aims to propel the development and application of foundation models across various disciplines by providing academic researchers with access to advanced AI models through Azure AI Services. The chatbot developed by Oh and her team is designed to assist English as a Foreign Language (EFL) students with their essay writing, offering guidance without writing the essays for them. Throughout a semester, 213 EFL students interacted with the chatbot, engaging with it as an intelligent peer and utilizing its feedback to refine their essays. The success of the project underscores the significant potential of generative AI in education, particularly in fostering a deeper understanding of problem-solving processes among students. The impact of AI extends beyond the healthcare sector, permeating major domestic industries such as semiconductors, education, batteries, finance, telecommunications, manufacturing, and content. As AI continues to drive scientific innovation and socioeconomic transformation, Korea is gearing up to reap its benefits to the fullest. It is also important to note that the foundation of this transformation lies in the hyperscale cloud infrastructure. This infrastructure provides the scalability and flexibility required for rapid innovation and robust AI applications.

AI governance in Korea

In pace with the rapid technological advancements and transformation triggered by AI, Korea was one of the first countries in the world to release a set of widely applicable principles that highlight the importance of grounding and guiding the use and development of technology such as AI in the digital space. ‘The Digital Bill of Rights: Charter for the Values and Principles for a Digital Society of Mutual Prosperity’ lays out the guiding framework for Korea’s vision on AI and influences the country’s legislative, regulatory, and operational approaches to AI. Korea’s vision for AI, as articulated by its government, is rooted in three core principles: 1) Responsible AI, 2) Inclusive AI, and 3) Sustainable AI.

Microsoft’s own principles for AI closely align with the principles of the ‘Digital Bill of Rights.’ Our focus is on Responsible AI, which emphasizes transparency, fairness, ethics, and accountability, while considering societal impacts and privacy. We also aim for Inclusive AI, ensuring equitable access and benefits for everyone. Moreover, we are dedicated to Sustainable AI, utilizing it for the betterment of humanity and minimizing adverse effects.

Moreover, our commitment is also to empowering a vibrant and open market for AI to flourish. During the Mobile World Congress (MWC) in March this year, we unveiled our commitment to responsible AI through the announcement of our ‘AI Access Principles.’ The 11 principles fall under 3 major themes: 1) Providing access and support for AI developers who create models and applications, 2) Ensuring choice and fairness across the AI economy, and 3) Meeting our societal responsibilities.

The ‘AI Access Principles’ serve as our compass, guiding Microsoft’s role and responsibility as an AI innovator and market leader. Our principles signal a profound shift, they pledge Microsoft to unprecedented investments, robust business partnerships, and expansive programs aimed at fostering innovation and competition. We recognize that AI’s transformative potential extends beyond corporate boundaries; it touches lives globally. By articulating these principles, we commit to providing broad technology access empowering organizations and individuals worldwide to wield AI for the greater good. Our initiatives around the world, including substantial AI datacenter investments and skilling programs, underscore our dedication to translating these principles into action.

Road to 2024 Seoul AI Safety Summit

As we approach the 2024 Seoul Safety Summit in May, we reflect on the strides made in the field of AI and the importance of ensuring its responsible development and deployment. The Summit serves as a platform for thought leaders, innovators, and policymakers to come together and discuss the guiding principles that will shape the future of AI. We applaud the Korean government’s efforts to work together with the global community to develop AI safety governance and policies that are globally coherent and foster a safe but collaborative approach to AI. Korea has a unique opportunity to take full advantage for global leadership in AI Safety governance in the upcoming 2024 Seoul AI Safety Summit by imploring and inspiring other nations, particularly in the Asia-Pacific region, to come together and cooperate for international regulatory coherence, as the world seeks to ensure safer governance of AI. This would help to foster trust, collaboration, and innovation across borders and sectors, and enable Korea to become a hub in the Asia-Pacific region for AI safety governance.

Article 28 of the ‘Digital Bill of Rights,’ which implores nations and companies to come together and cooperate to create universal digital norms and mechanism, could be leveraged, and specified as guiding principles for AI safety governance of highly advanced AI, i.e., Frontier AI. For example, Governments, Companies, Civil Societies, and Academia should:

  1. Work together in support of one another to develop universally coherent AI safety standards. These standards would provide guidance and benchmarks for ensuring the quality, reliability, and security of AI systems and their outcomes.
  2. Increase investments in developing evaluations for highly capable AI and foster partnerships for sharing best practices around how to develop and conduct these evaluations. Evaluations are essential for assessing the performance, impact, and risks of AI systems and ensuring that they align with the intended goals and values. By sharing best practice, countries can learn from each other and improve their evaluation methods and frameworks.
  3. Work to develop and implement policies for the identification, assessment, and management of risks related to highly capable AI models.
  4. Work to enhance transparency concerning the capabilities and risks of AI models and systems, and the policies and practices to ensure safety.
  5. Continue to evaluate and improve internal governance policies to make AI safer and more transparent. This would include compliance monitoring, and clear delegation of related roles and responsibilities for internal governance on AI safety.

In closing…

I would like to highlight that Microsoft is honored to be working with the National Information Society Agency (NIA) in developing this paper and hope it serves to demonstrate our commitment to the core principles embedded in Korea’s ‘Digital Bill of Rights’ as well as Microsoft’s ‘AI Access Principles’ which echo and resonate deeply with one another.

This partnership and our continuing activities in the global AI domain emphasize Korea’s position as a visionary leader in influencing global norms in AI governance.

We invite you to explore this whitepaper — a roadmap toward mutual prosperity through responsible AI governance. This whitepaper provides a comprehensive exploration of Korea’s strides in AI governance, reflecting our ongoing efforts to promote responsible, inclusive, and sustainable technological innovation on a global scale.

Together, let us navigate the digital frontier, ensuring AI serves humanity’s best interests.

관련 게시물