Knowledge Communication Secures Japan's First Order for F5 AI Guardrails in Gen AI Security, Completes PoC Project
Knowledge Communication has completed Japan's first Proof of Concept (PoC) for an enterprise generative AI security and governance platform utilizing F5's 'F5 AI Guardrails'. They provided end-to-end support, including Red Team validation.
📋 Article Processing Timeline
- 📰 Published: April 24, 2026 at 18:07
- 🔍 Collected: April 24, 2026 at 09:31
- 🤖 AI Analyzed: April 25, 2026 at 04:14 (18h 42m after Collected)
Knowledge Communication Co., Ltd. (Headquarters: Ichikawa City, Chiba Prefecture; Representative Director: Akira Okuzawa; hereinafter 'the Company') has completed a PoC (Proof of Concept) regarding the support for building a generative AI governance and security platform in a major enterprise company, utilizing the AI security solution 'F5 AI Guardrails' provided by F5 Networks Japan LLC. This project marks the first domestic order for F5 AI Guardrails (according to the Company's research).
In this PoC, covering both internal business use and external services, we consistently executed the organization of requirements based on the customer's existing guidelines, the examination of implementation architecture, and verification including Red Teaming (*). We confirmed the direction of an operational model that balances safety and business convenience, clarifying design issues for full-scale implementation.
* Red Team: A methodology to verify the system's resilience against anticipated attack patterns and inappropriate usage.
Security Challenges in Utilizing Generative AI
The scope of generative AI utilization in enterprises is rapidly expanding, from streamlining internal operations to sophisticated interactions at customer touchpoints. On the other hand, as the utilizing departments and usage forms expand, AI-specific risks that cannot be adequately addressed by conventional information security measures are emerging, such as prompt injection, unintentional leakage of confidential information, and inappropriate output generation.
In response to these challenges, rather than just individual risk countermeasures, a mechanism of control and protection is required to continuously expand the use of generative AI throughout the entire organization. This project tackles this issue head-on.
Overview of the Initiative
In this project, we provided support through the following steps:
Phase
Content
Requirements Definition
Defined the control and protection requirements necessary for generative AI utilization based on the customer's existing security guidelines
Architecture Design
Examined protection configurations suitable for each usage form: internal and external use
Test Planning & Execution
Formulated verification scenarios and executed various tests
Red Team Verification
Confirmed the comprehensiveness of protection functions through automated tests simulating attack patterns and inappropriate usage
Operational Model Organization
Organized the direction of operational design and remaining issues with an eye toward full-scale implementation
Main Outcomes of the PoC
Through this PoC, the following outcomes were confirmed:
Systematic organization of security and governance requirements for generative AI based on customer guidelines
Verification of the feasibility of protection architectures corresponding to usage forms (internal/external)
Confirmation of the effectiveness of protection functions through Red Team verification
Clarification of design issues and response policies toward full-scale implementation
Based on these results, we have organized the direction of an operational model premised on safety and business convenience, and are continuing discussions toward the next phase.
Reason the Company was Selected
The Company's strength lies in its ability to provide one-stop support from guideline formulation and introduction consulting for generative AI utilization to implementation, verification, and operational design. Our structure to provide hands-on support not only for organizing upstream policies but also for creating specific mechanisms with an eye toward actual operations was highly evaluated, leading to the award of this project.
Future Developments
The Company will continue to promote the safe and sustainable utilization of generative AI in enterprises through support utilizing F5 AI Guardrails. Specifically, we will strengthen support in the following areas:
Design and construction support for security and governance platforms
AI governance design and operational support
Test plan formulation and Red Team verification
Integration with the Knowledge Com AI Platform
Endorsement from F5 Networks Japan LLC
F5 Networks Japan LLC welcomes this initiative by Knowledge Communication utilizing F5 AI Guardrails and the completion of the PoC at a major enterprise company.
As the use of generative AI expands from internal operations to customer touchpoints, the importance of control and protection to balance safety and business convenience is increasing more than ever. F5 AI Guardrails supports protection and governance in generative AI utilization, providing a foundation for enterprises to safely promote AI usage.
Knowledge Communication significantly contributed to the advancement of this project by providing consistent support, from requirement organization based on the customer's existing guidelines to the examination of implementation architecture, test planning, verification, and evaluation including Red Teaming.
F5 will continue to contribute to the promotion of safe and sustainable generative AI utilization in Japanese enterprises through collaboration with Knowledge Communication.
Masanori Kimura, Country Manager, F5 Networks Japan LLC
In this PoC, covering both internal business use and external services, we consistently executed the organization of requirements based on the customer's existing guidelines, the examination of implementation architecture, and verification including Red Teaming (*). We confirmed the direction of an operational model that balances safety and business convenience, clarifying design issues for full-scale implementation.
* Red Team: A methodology to verify the system's resilience against anticipated attack patterns and inappropriate usage.
Security Challenges in Utilizing Generative AI
The scope of generative AI utilization in enterprises is rapidly expanding, from streamlining internal operations to sophisticated interactions at customer touchpoints. On the other hand, as the utilizing departments and usage forms expand, AI-specific risks that cannot be adequately addressed by conventional information security measures are emerging, such as prompt injection, unintentional leakage of confidential information, and inappropriate output generation.
In response to these challenges, rather than just individual risk countermeasures, a mechanism of control and protection is required to continuously expand the use of generative AI throughout the entire organization. This project tackles this issue head-on.
Overview of the Initiative
In this project, we provided support through the following steps:
Phase
Content
Requirements Definition
Defined the control and protection requirements necessary for generative AI utilization based on the customer's existing security guidelines
Architecture Design
Examined protection configurations suitable for each usage form: internal and external use
Test Planning & Execution
Formulated verification scenarios and executed various tests
Red Team Verification
Confirmed the comprehensiveness of protection functions through automated tests simulating attack patterns and inappropriate usage
Operational Model Organization
Organized the direction of operational design and remaining issues with an eye toward full-scale implementation
Main Outcomes of the PoC
Through this PoC, the following outcomes were confirmed:
Systematic organization of security and governance requirements for generative AI based on customer guidelines
Verification of the feasibility of protection architectures corresponding to usage forms (internal/external)
Confirmation of the effectiveness of protection functions through Red Team verification
Clarification of design issues and response policies toward full-scale implementation
Based on these results, we have organized the direction of an operational model premised on safety and business convenience, and are continuing discussions toward the next phase.
Reason the Company was Selected
The Company's strength lies in its ability to provide one-stop support from guideline formulation and introduction consulting for generative AI utilization to implementation, verification, and operational design. Our structure to provide hands-on support not only for organizing upstream policies but also for creating specific mechanisms with an eye toward actual operations was highly evaluated, leading to the award of this project.
Future Developments
The Company will continue to promote the safe and sustainable utilization of generative AI in enterprises through support utilizing F5 AI Guardrails. Specifically, we will strengthen support in the following areas:
Design and construction support for security and governance platforms
AI governance design and operational support
Test plan formulation and Red Team verification
Integration with the Knowledge Com AI Platform
Endorsement from F5 Networks Japan LLC
F5 Networks Japan LLC welcomes this initiative by Knowledge Communication utilizing F5 AI Guardrails and the completion of the PoC at a major enterprise company.
As the use of generative AI expands from internal operations to customer touchpoints, the importance of control and protection to balance safety and business convenience is increasing more than ever. F5 AI Guardrails supports protection and governance in generative AI utilization, providing a foundation for enterprises to safely promote AI usage.
Knowledge Communication significantly contributed to the advancement of this project by providing consistent support, from requirement organization based on the customer's existing guidelines to the examination of implementation architecture, test planning, verification, and evaluation including Red Teaming.
F5 will continue to contribute to the promotion of safe and sustainable generative AI utilization in Japanese enterprises through collaboration with Knowledge Communication.
Masanori Kimura, Country Manager, F5 Networks Japan LLC