AX-RAY & K-MYTHOS: Inside Korea's Consortium-Built Security-Specialized AI Foundation Model
AX-RAY & K-MYTHOS: Inside Korea's Consortium-Built Security-Specialized AI Foundation Model TL;DR: A 33-organization Korean consortium led by Naver Cloud is building K-MYTHOS ("K-미토스"), a security-domain foundation model that fuses threat intelligence from malware, dark web, network, and cloud sources into a dual-model attack/defense architecture. VIDRAFT contributes AI safety diagnostics and…
A consortium of 33 organizations in South Korea, led by Naver Cloud, is developing K-MYTHOS, a security-focused AI foundation model. This model combines threat intelligence from various sources, including malware, dark web, network, and cloud data, to create a dual-model attack/defense architecture.
The consortium includes security vendors, universities, and government institutions, each contributing their domain-specific threat data. The model is designed to be deployed across various security products, such as email gateways, network isolation appliances, and Security Operations Center (SOC) platforms. It has been validated in seven industry sectors.
The project's AI safety is handled by VIDRAFT, which uses its AX-RAY tool to perform a structured 117-point safety evaluation. This tool covers various inspection checklist items, ensuring the AI systems meet safety requirements before deployment. However, no quantitative benchmark scores have been published yet. AX-RAY is primarily used for evaluating the model outputs and deployment configuration, rather than the training data.
As of now, public access to K-MYTHOS and AX-RAY is not available, with the project still in the research and development phase.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.