日本語 
AGEST has started a proof of concept for AGEST ZERO SHIELD, a cybersecurity platform developed with US-based Lazarus that uses autonomous AI to strengthen enterprise security. The PoC, beginning at the end of July 2026, is the next step toward commercializing the platform, with its first zero-day vulnerability hunting capabilities planned for release in September.
The launch comes as AI is changing both sides of cybersecurity. While attackers are using more advanced AI models, security teams continue to deal with rising alert volumes, data sovereignty concerns, unpredictable usage costs, and restrictions around accessing frontier AI models. These challenges have made it harder for organizations to defend systems quickly and consistently.
AGEST ZERO SHIELD is designed to tackle those issues by running in on premises or closed environments, allowing companies to keep sensitive source code inside their own infrastructure. Instead of a token-based pricing model, the platform will use a fixed annual license. AGEST also plans to deliver regular updates to its cybersecurity focused AI model.
Also Read: AI-Powered SOCs: How Japanese Security Teams Are Redefining Threat Response
According to the company, the platform combines proprietary technology with tuned language models to improve vulnerability discovery, including identifying security flaws that traditional approaches may miss. The ongoing PoC will be carried out with partner companies, mainly large Japanese system integrators, to validate the platform in production-like environments before its commercial rollout.
Applications for the initial PoC have already closed due to strong interest. AGEST also plans to expand the platform with features such as live environment monitoring, security log analysis, autonomous AI-driven penetration testing, and offensive cyber control capabilities.
English