SANDS Lab, 42Maru and LG Uplus join forces to build a cybersecurity LLM with generative AI

SANDS Lab, 42Maru and LG Uplus join forces to build a cybersecurity LLM with generative AI

• First intercompany collaboration to build a cybersecurity-specific LLM
• A polished LLM expected to combine each company’s AI technology and R&D expertise
• Shared commitment to an innovative AI security environment using generative AI

Photo. SANDS Lab CEO Kim Ki-hong, 42Maru CEO Kim Dong-hwan and LG Uplus AI technology lead Han Young-seop (from left) at the agreement ceremony at LG Uplus’s Magok office.

SANDS Lab (KOSDAQ 411080, CEO Kim Ki-hong) announced on the 4th that it will jointly develop large language model (LLM) technology to strengthen cybersecurity with 42Maru (CEO Kim Dong-hwan) and LG Uplus (KOSPI 032640, CEO Hwang Hyeon-sik).

The ceremony at LG Uplus’s Magok office was attended by AI technology lead Han Young-seop, 42Maru CEO Kim Dong-hwan and SANDS Lab CEO Kim Ki-hong.

The three companies agreed to build LLMs specialized in four core cybersecurity areas—security monitoring, threat intelligence, anomaly analysis and vulnerability identification—and establish a deployment system. Once commercialized, businesses will be able to use and control LLMs internally while preventing information leakage risks.

LG Uplus will handle LLM development and safeguards for customer information. 42Maru will develop foundational technologies including cybersecurity and crime-prevention-specific LLM modeling and RAG, which accurately retrieves the specific information AI needs from large data volumes. SANDS Lab will contribute dataset development capabilities for operation within corporate infrastructure.

SANDS Lab is an AI security leader applying generative AI and training datasets across its services and solutions. It recently attracted attention by launching CTX for GPT in the GPT Store, making specialized intelligence accessible through natural language, and developing SANDY, a cybersecurity-specific small language model for internal corporate infrastructure. It is rapidly advancing its AI cybersecurity business.

Generative AI startup 42Maru is developing LLM42, a domain-specific language model for specialized industries, to address hallucination, security and cost issues in enterprise LLM adoption. Its deep learning-based reading comprehension solution MRC42 and RAG42 improve answer reliability; enterprise private mode protects internal and customer data; and commercial lightweight versions can dramatically reduce the cost of adopting and operating large-scale AI.

LG Uplus plans to launch its ixi-GEN LLM this year with LG AI Research and introduce an AI Agent, an evolution of chatbots, in the first half. It aims to use AI to deliver differentiated customer value, create new growth drivers and strengthen digital transformation capabilities.

SANDS Lab expects the agreement to significantly support commercialization of cybersecurity LLMs without information leakage risk, and anticipates synergies from jointly developing AI technologies to address cybersecurity and cybercrime.

CEO Kim Ki-hong said, “Combining specialists with strengths in cybersecurity and AI will maximize synergy. We will bring together our unique datasets and security expertise with 42Maru and LG Uplus’s advanced AI to create an innovative AI security environment.”

42Maru CEO Kim Dong-hwan said, “LLMs can enable proactive responses to various cybercrimes. We expect this cooperation to raise the level of Korean security technology.”

LG Uplus’s Han Young-seop said, “We will combine SANDS Lab’s security capabilities and 42Maru’s AI to build safeguards that protect customer information from cyber threats. Customer privacy is our top priority as we create a safer digital environment.”

SANDS Lab also plans to make its newly launched MNX NDR solution a growth driver this year. It will release a new AI-enhanced version that automates proactive threat response and is pursuing Public Procurement Service registration.