Cyberdigm Jaemin Lee Team Leader

"Generative AI has become a new channel for data leakage with evolving attacks. a series of incidents in which companies core confidential data leaked out to external models surfaced, and ransomware still targets endpoints and data scattered across the organization. At its root, the problem is that data remains dispersed across devices and every locations, left exposed as an ""attack surface."
"This session starts from recent major security incidents and the current state of enterprise AI adoption, examines the limits of conventional security, and proposes an approach that eliminates the attack surface. The starting point is document centralization. By managing dispersed assets centrally, internal leakage channels disappear, allowing both internal and external attacks to be blocked proactively, and enabling systematic asset management built on permissions, access control, and audit logs. Going further, when AI is connected only to verified, secure assets validated on-premises, a trustworthy AI Data Platform can be built without any data ever leaving the organization. Real customer cases will illustrate how assets secured through document centralization are integrated with and actively used by in-house AI.
""In the age of AI, autonomous security begins with document centralization."" In short, this session presents the most practical path to blocking threats proactively while using AI safely."