The Ministry of Science and ICT's 2026 work report placed 'AI Top 3 Nation' at the forefront. Large-scale GPU procurement, National AI Computing Center, independent foundation models, AX projects, and the AI Basic Act — 'AI' has become the central language of national operations.
However, from META-X's perspective, the more important question lies elsewhere. AI is just one engine constituting virtual convergence. What changes real industries, cities, and public services is not AI alone but the combination of virtual (digital) and reality (physical) — that is, virtual convergence.
Yet in this work report, virtual convergence is not a target industry nor an independent strategy but has been decomposed as a sub-function of 'AI Transformation (AX).' This is not simply a terminology change but shows that the coordinates of national technology strategy are moving from 'spatial transformation centered on metaverse, XR, and digital twin' to 'efficiency transformation centered on AI.'
'Virtual Convergence' Has Not Disappeared but Been 'Dismantled into AX'
Looking at the work report's regional and industrial strategies, functions that virtual convergence handled are distributed across multiple items.
- Establishing Physical AI diffusion strategy: The idea of applying AI to the physical world in manufacturing, logistics, and shipbuilding
- AI-RAN, 6G, intelligent base stations: Expanding precise control and demonstration through ultra-low latency, high-capacity networks
- Regional AX Project (total 3.1 trillion won): Large-scale demonstration applying AI transformation to region-specific industries such as manufacturing, biotech, and energy
- AI Government: Spreading 'work automation' utilizing AI from a common base for reports and press releases
The common point of this configuration is 'efficiency of real industries' rather than 'expansion of space (virtual world).' Digital twins, XR, and spatial computing are not visibly named but have implicitly been converted to prerequisite infrastructure for demonstration, control, and simulation. That is, virtual convergence has lost its 'narrative as industry' and only its 'function as a tool' remains.
The Core of Virtual Convergence Is Not 'AI' but 'Space, Data, and Control'
Virtual convergence was originally defined by the following combination.
- Space: XR/spatial computing, digital humans and UIs, combination of field and remote
- Model: Digital twin (virtual replication of processes, cities, healthcare, disasters)
- Data: Real-time collection and refinement of sensor, video, IoT, operational data
- Control: Controlling robots, equipment, mobile objects, and infrastructure through 'feedback loops'
- Network: Ultra-low latency, high reliability (5G SA→6G, AI-RAN)
- AI: Prediction, optimization, autonomization (ultimately 'one component')
What the work report emphasized is GPU, models, and AI utilization. But from a virtual convergence perspective, what is truly important is how the nation designs the loop of 'virtualizing reality (digital twin), verifying in virtual, and controlling reality again.'
This work report does not integrate and explain this loop under the name 'virtual convergence.' Instead it pursues it split into multiple pieces called AX, Physical AI, AI network, and regional demonstration. When policy is dispersed, even if technology is introduced, standards, linkage, and platform strategy tend to become gaps.
Why the 'Virtual Convergence Gap' Is Dangerous: Repetition of Demonstration Without Standards
Virtual convergence failures usually come not from 'technology insufficiency' but from connectivity insufficiency.
Industry-specific digital twins are built separately, XR/field UIs have different platforms and are incompatible, data is locked per institution and company, and control fails to expand due to safety and accountability issues.
The regional AX project (4 major zones) that the work report is strongly pushing is powerful in scale. But from a 'virtual convergence' perspective the question becomes sharper.
Will these demonstrations ultimately be bound by a single space, data, and control standard, or will they end as the sum of zonal pilots?
For virtual convergence to become national competitiveness, not individual demonstrations but 'reference architecture' replicable across industries is needed. That language is weak in this work report.
Core Diagnosis from META-X Perspective: 'AI Nation' Is Not 'Virtual Convergence Nation'
The government also announces plans for AI Basic Society, AI Basic Act, and AI Safety Ecosystem. This is the frame of 'a nation that uses AI safely.'
However, virtual convergence does not hold together with AI safety alone. Virtual convergence moves reality.
In healthcare, manufacturing, robotics, transportation, and disaster response, virtual convergence leads to physical results (accidents/liability/insurance/regulation). Therefore a virtual convergence national strategy needs (1) safety, (2) standards, (3) accountability systems, (4) data governance, and (5) demonstration-diffusion paths together.
This work report emphasizes 'safety' through the AI Basic Act and security strengthening, but the virtual convergence perspective's standards, accountability, and diffusion framework are hidden between AI and AX pieces. This is what META-X sees as the largest 'policy gap.'
The Name of Virtual Convergence Must Be Brought Out Again
This work report made the following choices.
- Elevated AI to national goals (GPU, models, budget, law)
- Removed virtual convergence from the 'name' of national strategy; instead pursuing through decomposition into AX, Physical AI, network, and regional demonstration
In this structure the most important question is this.
Can Korea secure leadership in 'virtual convergence transformation (spatial transformation)' just by growing AI big?
From META-X's perspective the answer is simple. AI is part of virtual convergence. And if the nation truly wants the next industry, now it is time to place virtual convergence's 'integrated strategy' back on document rather than AI.
