Artificial Intelligence (AI)
Trend Cybertron: Full Platform or Open-Source?
Previously exclusive to Trend Vision One customers, select Trend Cybertron models, datasets and agents are now available via open-source. Build advanced security solutions and join us in developing the next generation of AI security technology.
Save to Folio
Trend Cybertron: Trend Vision One “Cyber brain” now reshaping open-source AI Security
Trend Micro is excited to introduce Trend Cybertron — a groundbreaking advancement that is transforming cybersecurity in an increasingly AI-driven world.
While Trend Vision One customers benefit from the fully integrated Trend Cybertron "cyber brain," Trend Micro has now made select components of Trend Cybertron available as open-source. These offerings include cybersecurity-focused LLMs, extensive training datasets, and practical tools like the Cloud Risk Assessment AI Agent.
For developers and security professionals, this presents an unprecedented opportunity to build upon decades of cybersecurity expertise. Whether enhancing existing security tools, creating custom applications, or conducting research, organizations can now leverage models that have demonstrated significant improvements across security benchmarks. By joining this growing community, participants contribute to a collaborative ecosystem actively shaping how AI and cybersecurity intersect.
What is Trend Cybertron within the Trend Vision One platform?
Trend Cybertron, the industry's first proactive cybersecurity AI is fully integrated within the Trend Vision One platform designed for proactive detection and response. It’s a collection of specialized cybersecurity LLM models, datasets, machine learning technologies, natural language processing, and AI agents – combining comprehensive security expertise with our Trend Micro global threat intelligence and vulnerability research. This technology is the result of 20 years of AI investment, built on 35 years of industry-leading threat intelligence.
Through its integration with Trend Vision One, Trend Cybertron delivers a unified, intelligent security solution that not only protects against current threats but actively anticipates and prevents future security challenges. This forward-looking approach, combined with its comprehensive coverage across email, network, endpoint, and AI security domains, makes it an essential component of every modern enterprise security architecture.
This coordinated intelligence system represents Trend Micro's commitment to leveraging cutting-edge AI technology to provide comprehensive, proactive security solutions that protect organizations against both current and emerging threats in an increasingly complex digital landscape.
For Trend Vision One customers, Trend Cybertron is fully integrated into the platform and empowers customers to seamlessly experience the benefits within a unified cybersecurity platform.
How is Trend Cybertron different within open-source?
At Trend Micro, we believe in the power of community. That’s why we’re excited to share select models, datasets, and agents of Trend Cybertron with the open-source community. By fostering collaboration and innovation in AI-driven cybersecurity, we invite you to be a part of something bigger.
The following components of Trend Cybertron Open-Source are available through three main channels:
- NVIDIA NIM Catalog (https://build.nvidia.com/):
- Fine-tuned model using Llama 3.1 and optimized for deployment with NVIDIA NIM inference microservices
- Includes an 8-billion-parameter AI model with plans for a more advanced 70-billion parameter version
- Features multiple AI agent blueprints for seamless integration and security task automation
- Enables organizations to enhance security posture, reduce alert overload, save developer time, and extract greater value from existing security sensors
- Hugging Face (https://huggingface.co/trend-cybertron): Limited datasets and models from Trend Cybertron are available as part of the Primus collection and includes:
Models:
- trend-cybertron/Llama-Primus-Base: A text generation model based on Llama-3.1-8B-Instruct, continually pre-trained on 2.77B tokens of cybersecurity text, achieving a 15.88% improvement across cybersecurity benchmarks
- trend-cybertron/Llama-Primus-Merged: An instruction-following text generation model maintaining capabilities of Llama-3.1-8B-Instruct while achieving 14.84% improvement across cybersecurity benchmarks
- trend-cybertron/Llama-Primus-Reasoning: A text generation model distilled on reasoning and reflection data for cybersecurity tasks, where in testing achieved a 10% improvement on CISSP (Certified Information Systems Security Professional) practice exams versus humans.
Datasets:
- trend-cybertron/Primus-Seed: High-quality cybersecurity texts manually collected from reputable sources like Wikipedia, MITRE, cybersecurity company websites, and CTI
- trend-cybertron/Primus-FineWeb: 2.57B tokens of cybersecurity texts filtered from FineWeb
- trend-cybertron/Primus-Instruct: Approximately 1K QA pairs covering common cybersecurity business scenarios
- trend-cybertron/Primus-Reasoning: Reasoning and reflection data about cybersecurity tasks for distilled training.
- GitHub (https://github.com/trendmicro/cloud-risk-assessment-agent):
- Trend Cybertron - Cloud Risk Assessment AI Agent: An open-source project that scans source code, containers, Kubernetes clusters, and AWS infrastructure to detect CVE vulnerabilities, misconfigurations, and security risks
- Provides consolidated, prioritized risk assessment reports with actionable recommendations and remediation code snippets
- Features an interactive chat interface for investigating security issues
- Designed for Gen-AI developers, cloud infrastructure operators, and security teams
Read more details in this comprehensive whitepaper published by the Trend Micro AI Lab team on Cornell University arXiv portal about the Primus collection of open-source Datasets for Cybersecurity LLM Training.
Typical developer use cases for Trend Cybertron Open-Source
- Enhancing existing security tools with advanced AI capabilities
- Building custom security applications with specialized cybersecurity models
- Training and education in cybersecurity AI
- Research and development of new security techniques
- Integration into DevSecOps pipelines for automated security assessment
Trend Micro enables developers, data scientists, and security professionals to leverage decades of cybersecurity expertise while contributing to the advancement of the field through collaborative innovation in this increasingly complex digital landscape.
Why Developers Should Join the Community
If you're working in AI application development, there are compelling reasons to engage with Trend Cybertron Open-Source:
- Cybersecurity Expertise You Can Trust: Build with confidence on Trend Micro's battle-tested security foundation. Our open-source cybersecurity LLMs, datasets, and agents are backed by decades of threat intelligence experience and our proven track record protecting millions of endpoints worldwide. When you use our AI tools, you're leveraging security intelligence that's been validated in real-world environments.
- Access to Specialized Models: Leverage cybersecurity-focused LLMs that have demonstrated significant improvements across security benchmarks.
- Real-World Applications: From threat intelligence analysis to vulnerability assessment, malware detection, and incident response - these tools address practical security challenges.
- Collaborative Innovation: Join a community that's actively shaping the future of AI in cybersecurity, with contributions from both industry experts and independent researchers.
- Enterprise-Grade Foundation: Build on technology developed by a company with 35 years of security expertise and trusted by 20 of the top 25 Fortune Global 500 companies.
- Standards Influence: Contribute to a project that's helping shape international AI security standards and policies.
Trend Cybertron Open-Source components offer an excellent entry point for developers looking to strengthen their applications with robust security capabilities, while its fully integrated version within Trend Vision One offers a comprehensive enterprise solution for organizations seeking full protection.
The Team Behind the Innovation
Behind this groundbreaking technology is an exceptional team of Trend Micro innovators whose expertise spans cybersecurity, artificial intelligence, and software development. Their commitment to both technical excellence and open collaboration demonstrates Trend Micro's leadership in advancing the field of AI-powered cybersecurity.
Open Cybertron Product Management Leaders
- David Girard – Product Manager
- Fernando Cardoso – Product Manager
Open Cybertron Taskforce
- Jason Yueh – Sr Distinguished Technologist
- Stephanie Chou – Sr Manager (AI Labs)
- Patrick Lu - Project Lead
- Spark Tsao - Project Lead
- Kevin K Chang - Domain Expert
- Willy Kao - Domain Expert
- Hank Yang – LLM/NIM Inference and model evaluator
- Addison Kao – LLM/NIM Inference and model evaluator
Trend Micro AI Lab Team
- Liam Huang - AI Research Lead
- Yao-Ching Yu - LLM Training Expert
- Dennis Tsai – AI Researcher
- James Chiang - Data Engineering Lead
Join the Movement
Whether you're building the next generation of secure AI applications, enhancing your organization's security posture, or simply passionate about advancing cybersecurity through open collaboration, Trend Cybertron Open-Source provides a powerful foundation to build upon.
By contributing to this community, you're not just improving your own projects -- you're helping shape the future of cybersecurity in an AI-driven world.
Check out the resources on GitHub, Hugging Face, and NVIDIA's NIM Catalog to get started with Trend Cybertron today!