Beyond Intelligence to Safety: The Ultimate Guide to 'External AI Guardrails' in the AI Era

Time: 1:55 - 2:25 PM JST
Language: KO

Even as AI becomes more advanced, security threats do not disappear. Intelligence does not guarantee safety. Instructions given to the model—that is, the system prompt—cannot fully prevent these threats on their own. We propose adopting 'external AI guardrails' that go beyond the limitations of system prompts.

External guardrails enforce safety policies outside the model, improving cost efficiency and operational control while building multilayered defense. In this presentation, we introduce five key guardrails.

Text Moderation
Prompt Injection Detection
PII Filter
Topic Control
Hallucination Detection

If you want to build trustworthy AI services, consider adopting external guardrails that protect what lies outside the model.

View Materials

Jongwoo Han

LINE PlusApplied ML Dev

Ph.D., Korea University (2011). Worked at LG Electronics (2011–2020) and Intel Korea (2020–2021). Currently Lead, Applied ML Dev at LINE Plus (2022–present).

Hyukjae Jang

LINE PlusApplied ML Dev

He is a software engineer at LINE Plus who brings AI Guardrails and on-device ML into real services.
Recently, he has been designing, training, and deploying multilingual AI safety systems centered on Topic Control, while also building benchmarks and automated evaluation pipelines himself.
His experience spans LINE Android client development, federated learning, on-device AI, and generative AI, and these days he is especially passionate about turning “AI that works in demos” into “AI that holds up in products.".

Sooah Lee

LINE PlusApplied ML Dev

AI Software Engineer at Samsung Electronics, specializing in Coding Agent systems, LLM serving platforms, and GPU inference optimization. Previously worked at Intel Korea on Datumaro, data pruning, multimodal dataset management, and model compression. Currently AI Software Engineer at LINE Plus, driving hallucination detection for the AI Guardrail project through RAG-, NLI-, and LLM-based verification architectures, evaluation frameworks, and operational pipelines.

Younghyun Kim

LINE PlusSecurity R&D

Research Engineer at Samsung SDS (2018–2021), specializing in Privacy-Preserving Machine Learning with Homomorphic Encryption and Differential Privacy. Currently Security Research Engineer at LINE Plus (2021–present), driving security R&D across FIDO2, Device Attestation, Privacy Enhancing Technologies, and AI Guardrails.

All Sessions