Blockchain technology has evolved from a nascent concept aimed at securing decentralized digital assets to a powerful framework for enhancing security in diverse domains.
This talk explores the dual trajectory of blockchain’s role in security: from addressing inherent vulnerabilities within blockchain protocols to using blockchain as a security-enabling infrastructure.
First, we examine key advancements in securing blockchain systems, including cryptographic safeguards, and how AI can contribute to detecting threats and enhancing overall security.
Then, we shift focus to emerging applications where blockchain serves as a foundational layer for implementing secure data sharing and privacy-preserving protocols in AI frameworks such as federated learning.
By highlighting both aspects, this talk underscores the transformative potential of blockchain as both a subject and a tool for advancing security in distributed systems.

As attackers change their strategies over time, the threats in cybersecurity continually evolve. Artificial Intelligence (AI) enhances cybersecurity by autonomously detecting and responding to threats, hence augmenting the capacity to defend against various cyberattacks in a dynamic manner. This talk emphasizes the benefits of AI-assisted systems that effectively manage complex cyberattacks while minimizing human errors. I also address the risks associated with adversarial AI and the ethical implications for data privacy. The talk finishes by forecasting the necessity of integrating AI into cybersecurity systems as a pivotal element for ensuring a secure future Internet.

With the expanding adoption of large language models (LLMs) in the public sector, ensuring security and data sovereignty requires a domestically controlled, network-separated closed infrastructure. Rather than one-off development, a sustainable operational ecosystem must be established.

[Directions for Building a Sustainable Foundation Model]
Public-private collaboration is essential for developing domain-specific LLMs tailored to high-risk and low-profit areas (e.g., healthcare, legal sectors), based on open-source LLMs such as LLaMA, DeepSeek, LG EXAONE, and Naver HyperCLOVA X.

[Five Strategic Frameworks for Public LLM Development]
1. **Model/Architecture**: Build internal capabilities and flexible architecture to incorporate cutting-edge technologies.
2. **Data**: Establish a high-quality public data lake through refined integration.
3. **Infrastructure**: Expand secure infrastructures based on GPU, on-premise, and network separation.
4. **Talent & Organization**: Train in-house experts and build a collaborative public-private system.
5. **Policy & Regulation**: Institutionalize legal frameworks and operational standards for research data.

The research focuses on strategies for securing autonomy, sustainability, and security in public-sector AI — paving the way toward national AI sovereignty and enhanced global competitiveness.

Clumit Security is an autonomous cybersecurity platform developed by ClumL, integrating proprietary AI clustering and large language model (LLM) orchestration technologies.
In this presentation, we introduce the core principles and real-world applications of ClumL’s AI clustering engine, which analyzes large-scale network traffic in real time. Unlike traditional systems that rely on predefined rules or signatures, our technology effectively detects emerging, mutated, and stealth threats.
Moreover, Clumit Security leverages LLMs to automatically interpret detected events and generate human-readable reports—reducing repetitive manual tasks and enabling security professionals to focus on high-level strategic decisions.
By transcending the limitations of conventional security architectures, Clumit Security paves the way for a faster, smarter, and more precise era of autonomous cybersecurity.

The IT industry is evolving at an unprecedented pace—and at the heart of this transformation lies cloud technology. For today’s enterprises, moving beyond traditional infrastructure and adopting a cloud-native approach has become a strategic necessity.
POPCON HCI offers a seamless solution for this transition by enabling unified management of both VM-based systems and Kubernetes-based container environments within a single platform. This reduces operational complexity while allowing for gradual migration with scalability and resource optimization in mind.
Designed to maximize the value of existing IT assets, POPCON HCI empowers organizations to flexibly adopt new cloud-native technologies without abandoning prior investments.
As a comprehensive solution, POPCON HCI helps businesses overcome the limitations of legacy systems and shift efficiently and reliably into a modern, cloud-native infrastructure.

This session explores strategic approaches to implementing a national network security framework (N²SF) based on Zero Trust principles.
It introduces a methodology for classifying data by Confidentiality, Sensitivity, and Openness (C/S/O), and for assessing Zero Trust maturity to derive AS-IS and TO-BE models. The strategy aims to prevent redundant investments and mitigate policy conflicts between overlapping control mechanisms.
In particular, the session will present the “Zero Trust Overlay” approach—a practical framework for introducing Zero Trust architecture effectively across complex government or enterprise environments.

The Security Operations Center (SOC)—the heart of cybersecurity—is evolving. At the core of this transformation lie AI technologies, the N²SF (National Network Security Framework), and the need for skilled professionals who can understand and operate within this new paradigm.
This session explores the changes required across key domains to achieve successful SOC evolution. It also highlights practical strategies and real-world implementations adopted by Igloo Corporation to support this transition.
As security threats grow more complex, the future of SOC demands not only technological advancement but also strategic alignment and workforce readiness.

Many students today begin their journey into hacking through CTF (Capture The Flag) competitions, sharpening their skills by solving complex and challenging problems. However, a common question arises: ""How does this connect to my future career?"" ""What more should I do to grow as a professional hacker?"" In an era where technology evolves rapidly and AI is transforming the landscape—including cybersecurity—these concerns are more relevant than ever: ""How can I survive and thrive in the midst of such change?"" In this talk, the speaker shares personal experiences—first as a student who grew through CTFs, and now as a professor mentoring future hackers. The session aims to provide insights and practical advice on building a meaningful career in cybersecurity, navigating industry shifts, and finding your own path as a hacker in the age of AI.

Generative AI is driving innovation across various industries, including the field of cybersecurity. While it offers powerful tools for both offense and defense, it also introduces new and significant security challenges.

This talk explores the dual dimensions of security in the era of Generative AI:
1. **AI for Security** – the use of AI technologies to enhance cybersecurity capabilities, such as phishing detection, malware generation, code auditing, and log analysis.
2. **Security for AI** – addressing vulnerabilities in AI systems themselves, including prompt injection, data poisoning, and corresponding defense mechanisms like guardrails and robust alignment strategies.

By examining both how AI can be used to secure systems and how AI systems must be secured, this presentation offers a comprehensive overview of cybersecurity in the age of Generative AI.

With the rapid advancement of quantum computing, concerns are growing over the vulnerability of existing public-key cryptographic systems. To prepare for potential threats in the quantum era, researchers and institutions are actively exploring various countermeasures.

This talk introduces the fundamentals of quantum computing and recent developments in quantum software. It then examines the current state of post-quantum cryptography (PQC) standardization efforts aimed at securing systems against quantum-enabled attacks. Additionally, the session will explore emerging security technologies that leverage the unique characteristics of quantum mechanics.

Currently, more than 7,000 low Earth orbit (LEO) satellites are in operation, and the number is rapidly increasing each year with the expansion of networks like Starlink. Despite this growth, there is limited public knowledge about the nature of the data transmitted and received through these satellites.

This presentation briefly shares preliminary findings from our research team’s recent project on collecting and analyzing LEO satellite signals. While still in its early stages, we believe that large-scale satellite signal analysis will play a crucial role in shaping the future of space cybersecurity.