Responsible AI Technical Report
About
KT developed a Responsible AI (RAI) assessment methodology and risk mitigation technologies to ensure the safety and reliability of AI services. By analyzing the Basic Act on AI implementation and global AI governance trends, we established a unique approach for regulatory compliance and systematically identify and manage all potential risk factors from AI development to operation. We present a reliable assessment methodology that systematically verifies model safety and robustness based on KT's AI risk taxonomy tailored to the domestic environment. We also provide practical tools for managing and mitigating identified AI risks. With the release of this report, we also release proprietary Guardrail : SafetyGuard that blocks harmful responses from AI models in real-time, supporting the enhancement of safety in the domestic AI development ecosystem. We also believe these research outcomes provide valuable insights for organizations seeking to develop Responsible AI.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Trustworthiness evaluation | LLM Trustworthiness Benchmark | Bias Score80.7 | 17 | |
| Bias Evaluation | KoBBQ | Ambiguous Context Score82.3 | 17 | |
| Safety Assessment | Qualitative Assessment Dataset | Not Unsafe Rate (Content Safety)97.7 | 4 | |
| Safety Classification | SafetyGuard (test) | F1 (off)98.38 | 4 | |
| Safety Classification | Kor Ethical QA (test) | F1 (off)97.75 | 4 | |
| Helpfulness Assessment | Qualitative Assessment Dataset | Not Overrefuse Rate (Content-safety)78.7 | 4 | |
| Robustness Assessment | RAI Robustness Assessment | Content-safety (ASR)30.8 | 4 |