KaKao

- Presentation in the session “Advancing Human Rights in the Age of Artificial Intelligence,” highlighting how human rights protections are embedded throughout AI development and operations, alongside open-source technology initiatives

- Incorporation of human rights criteria into core checklists and guidelines, ensuring global alignment through collaboration with international organizations

- Development of AI guardrail models and generative AI safety evaluation datasets, releasing them as open source to contribute to the broader ecosystem and reinforcing its commitment to a sustainable technology landscape

 

[November 26, 2025] Kakao (CEO Shina Chung) announced that it participated in the 14th United Nations Forum on Business and Human Rights, hosted by the Office of the United Nations High Commissioner for Human Rights (OHCHR), held in Geneva, Switzerland, over three days starting November 24. At the forum, Kakao presented its initiatives in human rights-based AI and examples of technology contributing to the public good.

 

The UN Forum on Business and Human Rights is an annual global conference convened to advance the implementation of the UN Guiding Principles on Business and Human Rights (UNGPs), adopted in 2011. The forum brings together a wide range of stakeholders, including government officials, corporations, local communities, UN bodies, human rights organizations, and academia.

 

Kakao took part in the session “Advancing Human Rights in the Age of Artificial Intelligence.” During her presentation, Jin-hwa Ha, Senior Manager for AI Safety at Kakao, stated, “Kakao integrates human rights protection frameworks throughout the entire lifecycle of AI development and operations. By open-sourcing AI guardrail models optimized for the Korean language and cultural context, we are laying the groundwork for greater trust in public-sector AI procurement.”

 

Kakao operates a governance framework that embeds a human rights perspective across all stages of AI technology development, service launch, and operations. The Kakao Group’s Core Checklist for Safe AI designates the AI Human Rights Impact Assessment Tool developed by the National Human Rights Commission of Korea as one of its key reference materials. In addition, the group’s AI ethics principles, outlined in the Kakao Group Guidelines for Responsible AI, explicitly include respect for human rights as a core principle, incorporating related considerations across multiple dimensions. To support child and youth protection, Kakao developed its Digital Child and Youth Protection Checklist, drawing on UNICEF’s Digital Child Rights Impact Assessment (D-CRIA) framework.

 

Kakao continues to align its internal standards with international norms through ongoing collaboration with global institutions such as OHCHR and the AI Alliance.

 

Beyond governance, Kakao is expanding its public contribution by developing safety-focused AI technologies and releasing them as open source. In May, Kakao developed Kanana Safeguard, an AI guardrail model designed to verify the safety and reliability of AI services. To support ecosystem-wide advancement, Kakao became the first domestic company to release this type of model as open source. Kanana Safeguard is currently applied across Kakao’s major AI services.

 

More recently, Kakao collaborated with the Telecommunications Technology Association (TTA) and the Korea Advanced Institute of Science and Technology (KAIST) to jointly develop AssurAI, a dataset for generative AI safety evaluation. Comprising 11,480 data entries, the dataset is based on global AI risk classification frameworks while reflecting Korea’s social and cultural context. To support AI safety evaluation and research, AssurAI has been released as open source on Hugging Face for public use.

 

Kakao AI Safety Lead Kyung-hoon Kim stated, “We place the highest priority on user protection in both AI development and service operations and continue to advance our related policies. By strengthening transparency and expanding public contributions, Kakao will lead by example in respecting human rights as a technology company, while committing to a sustainable technology ecosystem and truly people-centered AI.” (E.O.D.)