Kaspersky Launches AI Protect in Asia Pacific to Secure Enterprise AI Infrastructure

Kaspersky is launching Kaspersky AI Protect (KAIP) across Asia Pacific, introducing a security solution designed to check large language model files and AI agents for malicious functionality.

The product targets risks associated with enterprise AI adoption, including threats introduced through externally sourced models, plugins, and agent files. According to Kaspersky, it also retains protection against conventional cyberthreats.

The launch focuses on organizations incorporating AI into business applications and workflows, where new software components can create additional security concerns.

In-house AI still needs security checks

Kaspersky says hosting an AI model internally does not remove the risks associated with materials imported from outside an organization.

“Many organizations still hold the misconception that keeping large language models in-house makes them safe. But that overlooks how malicious code can be smuggled in through model weights, AI plugins, and skill scripts sourced from outside the organization,” said Adrian Hia, managing director for Asia Pacific at Kaspersky.

The company positions KAIP as a way to inspect these components as part of existing security operations.

Its announcement highlights dedicated checks for LLM files and AI agents. Although it also discusses risks such as prompt injection, it does not detail a separate protection mechanism for every threat mentioned.

Integration with enterprise applications

According to Kaspersky, KAIP can integrate with web portals and applications, proxy servers, network-attached storage, and mail gateways.

It supports integration through HTTP and ICAP, with deployment options including a standalone service, a scalable cluster, or a Docker container.

The product is available for Linux environments and can be deployed in an individual container, Docker Swarm, Kubernetes, AWS EKS, and similar cloud environments.

These options are intended to accommodate different enterprise infrastructure setups. The release does not provide independent performance test results or deployment benchmarks.

Security across the AI lifecycle

Kaspersky’s stated approach covers security from model import to agent execution, with the aim of incorporating AI-related checks into established business security workflows.

“Security should not act as a barrier to AI advancement, but rather as an enabler that unlocks AI’s scalable value,” Hia said.

The company also says KAIP can become part of third-party hardware and software security offerings through its Kaspersky Technology Alliance ecosystem, with technical cooperation support for partners.

Responding to threats linked to AI adoption

Kaspersky reports recording more than 92,000 malicious attacks disguised as AI services globally in the first half of 2026. That figure is company-reported global data, rather than a count limited to Asia Pacific or the Philippines.

For enterprises, the launch highlights the need to examine the components entering an AI environment alongside the systems hosting them. KAIP is Kaspersky’s offering for adding those checks to enterprise infrastructure.

The announcement does not include Philippine pricing or licensing details. More information is available through Kaspersky’s official website.