Private AI has rebranded from zero-trust security firm Private Technology into a company focused on AI governance and control. The company said the shift from zero trust to AI governance is a natural evolution, not a move to follow a trend of adopting AI-related names.

Private Technology is reborn as "Private AI" and unveils a strategy to transform from security to AI governance and risk control, moving beyond zero trust to an AI operations platform.

Private AI executive director Beom-su Joo (주범수) said in a recent interview with a reporter, "We expanded into AI because our core technology itself fits well with AI governance." He said, "AI governance is not just a simple vision; we have already implemented it as products." He added, "Our goal is to raise the share of AI in revenue to more than half next year."

Since its founding in 2018, Private AI has focused on developing ZTNA (Zero Trust Network Access) technology that applies the zero-trust principle of "never trust, always verify" to network access control.

The company said its ZTNA capabilities can be applied beyond security to AI governance and control. It said it could offer related products in a short period after declaring its AI business because it reused existing technology rather than developing everything from scratch.

Joo said, "Privilege management, monitoring, and access and execution control are elements that make up AI governance." He said companies must fine-tune AI while protecting internal data and also support routing to external models. "That is how you can protect data and also build an environment for cost management," he said.

Private AI is approaching its AI governance and control platform from two main perspectives. One is detection, focusing on asset identification, classification by tier, tracking data movement paths and threat analysis. The other is control, supporting decisions and enforcement on whether packets are allowed to pass and whether agents are allowed to run.

Joo said, "Whether for people or AI agents, we already had know-how in access-rights management. We shifted direction to optimize that for AI." He said the company is focusing on controlling and managing whether AI agents maintain confidentiality when calling tools, linking to RAG (Retrieval-Augmented Generation) and communicating with each other. He added, "Private AI's strength is that it can directly control AI at the execution layer that takes place in networks and sessions. We also provide a kill switch function."

Private AI has set the financial sector and public institutions as its primary initial targets for its AI governance platform. For the public sector, it sees the rollout of the National Network Security Framework (N2SF), which focuses on easing network separation regulations, as an opportunity.

With the introduction of N2SF this year, public institutions' security environments have moved away from a uniform network separation structure, making it possible to publish less important information on the internet. N2SF classifies data importance into C (Classified), S (Sensitive) and O (Open), and allows O-grade data to be published online. Joo said N2SF has lowered barriers for public institutions to use external AI services, and he expects this will expand the AI governance solution market. "We are currently conducting proof-of-concept (PoC) projects with several public institutions," he said.

Private AI also has a roadmap to platform its AI execution control solution and provide it as software as a service (SaaS). It aims to use this to enter overseas markets. Joo said, "As network separation regulations have been removed, an atmosphere is forming to move to services. Monitoring and execution control solutions will be included in the platform."

The AI governance market is already a battle among many players. With major enterprise software and security firms as well as big cloud companies joining the race, it does not appear easy for smaller firms to lead the market. On this, Joo said, "Rather than doing everything alone, we will work with companies running AI transformation (AX) businesses, cover the governance part in AX projects, and expand our foothold."