Trend Micro Japan releases AI Security trap… "The core is in preemptive threat interception"

Japanese cybersecurity company Trend Micro has launched a security solution called “Trend Vision Meta AI Security Trap” aimed at protecting the overall environment of artificial intelligence, marking the beginning of a transformation in the AI security market landscape. This product will officially debut at the AWS re:Invent conference held in early December, with a design capable of identifying and responding to security threats throughout the entire process from AI model development to operation.

Trend Micro points out that enterprises generally lack visibility into the system decision-making processes, data processing methods, and the potential for malicious access when introducing AI technologies, while traditional endpoint, network, and cloud security solutions have become inadequate. In particular, the recent frequent occurrences of threats unique to AI, such as prompt injection, data poisoning, and output manipulation, represent blind spots that traditional systems struggle to detect.

In response to this trend, Trend Micro announced that it will address these issues by building an AI-exclusive security system capable of detecting internal risk factors of AI models and automatically enabling intelligent protective devices. This system integrates NVIDIA's BlueField-3 technology, enhancing security protection throughout the entire process from hardware acceleration to model deployment.

The Trend Vision Meta AI Security Suite includes an “AI Scanner” that continuously monitors AI model vulnerabilities and proactively addresses threats. Additionally, the dedicated AI security tools “AI Security Blueprint” and “Risk Insights” visualize the construction of an AI governance system, clarify internal regulatory requirements, and ensure compliance. Its core focus is on protecting confidential models throughout the entire AI development process and providing practical security implementation solutions.

The suite also includes various left-shift security solutions such as “Cloud Risk Management” for providing project-level risk visibility and “Container and Code Security Tools” for early assessment of security vulnerabilities in the coding phase. Notably, “File Security” integrates with NetApp storage systems to respond in real-time to malware and ransomware attacks in cloud storage.

Operational security in the AI environment has also been strengthened. The native integration of AWS logs to implement AI detection and response “intelligent SIEM”, as well as “zero trust security access” to control the scope of generative AI usage, are solutions focused on minimizing sensitive information leakage and shadow IT risks.

Trend Micro Chief Platform Officer Rachel Jin emphasized: “Uncontrolled innovation will bring unbearable risks to enterprises. In the current context of widespread AI application, our goal is to provide enterprises with a safety barrier based on security and trust, to support new growth.”

This release is an extension of the digital twin AI Security model jointly launched by Trend Micro and NVIDIA in July of this year, and is seen as an important component of a scalable AI risk management strategy in enterprise environments. As AI Security gradually becomes a core competitive advantage for enterprises, the industry is closely watching what kind of transformation Trend Micro's comprehensive response strategy will lead to.