Editorial
The Gist
- The metaverse didn’t fail — it misjudged its primary user. Early predictions assumed humans would migrate en masse into immersive 3D worlds, but adoption stalled because the value proposition depended on costly hardware and unrealistic behavior shifts.
- AI agents are becoming the true inhabitants of the metaverse. The same spatial infrastructure built for human immersion is now being repurposed as training, coordination, and transaction environments for autonomous AI systems operating at machine speed.
- The metaverse is evolving into the backbone of agent-to-agent commerce. As the Spatial Web takes shape, standardized protocols and semantic environments enable AI agents to negotiate, transact, and manage supply chains without human intervention.
It is easy to make predictions, and even easier to critique those that others made in the past. Case in point: Gartner predicted in 2022 that by 2026 (yes, that's this year), 25% of the population will spend nearly one hour per day in the metaverse.
The premise of this largescale adoption, which McKinsey predicted would generate up to $5 trillion in value by 2030, relied on a massive behavioral shift: that humans, represented by avatars, would migrate their social, professional and economic lives into 3D environments accessed via (at the time, and currently still) expensive VR and AR hardware.
We all know what happened: today, the metaverse is a "ghost town" with few human users, and (it can be assumed) many disappointed investors, entrepreneurs and others. Yet, despite the false (and costly) start, the metaverse may have its own second life (no pun intended).
While the consumer-facing metaverse never achieved mass adoption beyond gaming niches, parallel advancements in artificial intelligence began to repurpose the infrastructure built for human immersion into a metaverse for machines, namely those agentic AI systems we've all been hearing so much about. Thus, the primary users of this newer interaction are autonomous AI agents interacting within a Spatial Web governed by shared standards, where the "user" is software, and the "experience" is high-frequency, autonomous commerce.
Thus, the metaverse is still here, is ripe with purpose, and it no longer requires direct human interaction. Instead, the metaverse has found its audience in the Agent-to-Agent (A2A) machine economy. The complex 3D environments and spatial protocols originally designed for human sensory immersion are now serving as essential training grounds for autonomous AI agents, which require semantic, spatial and mathematical representations of the world (inherently provided by the metaverse) to navigate, negotiate, and transact effectively.
Table of Contents
- The Emergence of the Machine User
- Why AI Agents Need a Metaverse of Their Own
- Retailers' Transition to Agent-to-Agent Commerce
- Summary: Will the Metaverse Survive?
The Emergence of the Machine User
The critical realization for the industry has been that the primary resident of the metaverse will likely be the AI Agent, an autonomous system capable of perceiving its environment, reasoning about it, and taking actions to achieve specific goals without constant human intervention. Unlike a passive chatbot, an agent performs actions on behalf of humans, such as booking travel, negotiating supply chain contracts, operating robots in a warehouse, or managing retail inventory.
These agents that evolved from text-based LLMs to Embodied AI that consists of robots, drones and digital assistants, face a challenge with the context of the real, three-dimensional world. For instance, an AI trained solely on text knows that a chair is a noun often associated with sitting, but it does not understand the spatial physics of a chair, its weight, or its location in a room relative to a door.
The metaverse provides this missing layer of context. It offers a Spatial Web, a geolocated, physics-compliant layer of the internet, where agents can anchor their knowledge. Consequently, the metaverse is evolving into an invisible layer of the internet where agents interact, negotiate and transact at speeds incomprehensible to humans.
Related Article: Agentic AI and Marketing: The Death of the Traditional Funnel?
Why AI Agents Need a Metaverse of Their Own
The current web infrastructure is not exactly friendly to agents. Scraping websites' HTML is unreliable, since a minor UI change by a retailer can break a shopping bot. Current AI models, like the ones that power conversational AI chatbots, struggle because they often don't have a real-world foundation; they can "hallucinate" or provide incorrect information because they lack a physical sense of location or action.
Thus, an AI agent needs to understand objects, time and space to become truly autonomous. This is done through a process called active inference, constantly adjusting actions based on its predictions and what it senses. Without a defined, structured environment, however, this level of intelligence is impossible to achieve.
This is where the metaverse, or the Spatial Web, offers a solution. Think of it as giving the internet a coordinate system. Instead of being a collection of web pages with simple links, the Spatial Web is a network of defined spaces and locations. This means an AI agent knows that "Product A" is physically located at "Address X" at "Time T." This spatial grounding gives the AI the necessary context to understand the consequences of its actions, dramatically reducing errors and enabling the kind of reliable, autonomous behavior that businesses will demand from future intelligent agents.
Retailers' Transition to Agent-to-Agent Commerce
While the traditional B2C model isn't going anywhere anytime soon, we are seeing a steady rise in Agent-to-Agent (A2A) interactions, and significant investment of resources into building out an agentic economy. In this A2A model, a consumer's buying agent interacts with a business's selling agent to negotiate and transact without human intervention.
A2A commerce, utilizing the Spatial Web behaves similarly to ecommerce of today, yet with some meaningful differences.
A consumer instructs their agent: "Find me a pair of running shoes, size 10, under $150, that are highly rated for arch support and can be delivered by Friday." The agent does not "browse" Amazon. It queries the spatial web. At the same time, retailers deploy autonomous agents that monitor inventory, demand and competitor pricing in real-time. When queried by a Buying Agent, the Selling Agent instantly negotiates: "I have that shoe. Listed at $160, but I can offer it to you for $148 if you complete the transaction within 50 milliseconds".
Behind the scenes, standardized protocols like Model Context Protocol (MCP), and other Agent to Agent protocols, enable communication between a consumer and a retailer's agents.
Webinar
Jan
28
Ditch the Desk Phones: How Modern Teams Drive AI-First Communications
Find out how one team finally pulled the plug on a legacy phone system. And built something smarter.
Register
Webinar
Jan
29
Cut the Noise: Deploying AI That Actually Moves the Needle
Learn how to turn AI experimentation into concrete revenue operations.
Register
Webinar
On demand
Empowering Non-Profits: Smarter Crisis Communication and Community Engagement
The cost of miscommunication is measured in more than words. Learn to deliver outreach that's fast, clear and trusted.
Watch Now
Webinar
On demand
Roundtable: Turning Real-Time CX Signals into Business Results
Four big brands. One live, unscripted discussion on how modern CX teams move from dashboards to real impact.
Watch Now
Webinar
On demand
Unlock Connected Service: How to Forecast, Staff & Support Every Channel
Stop juggling tools. 73% of CX leaders say silos damage CX. Build a seamless service operation instead.
Watch Now
Webinar
On demand
Beyond Composability: How Modern Marketers Build Connected Experiences
Ready to launch campaigns faster, personalize smarter and prove your marketing ROI? Discover the power of a modern DXP.
Watch Now
How Agentic Commerce Works With the Metaverse
The metaverse is evolving from a consumer-facing environment into a machine-native infrastructure that supports autonomous AI agents, spatial reasoning and agent-to-agent commerce.
| Layer | What’s Changing | Why It Matters for Agentic Commerce |
|---|---|---|
| Machine-Only Metaverse Infrastructure | New internet standards extend beyond HTML to support 3D, physics-aware environments where AI agents, IoT devices and robots can interact. Key components include IEEE P2874 for spatial interoperability, HSML for object identity and traits, HSTP for permissions and transactions and OpenUSD to ensure consistent behavior across digital twins and virtual environments. | These standards give AI agents a shared, spatially grounded environment where they can reason about objects, locations and actions reliably, enabling autonomous interaction at machine speed without human mediation. |
| Retailers’ Agent-Ready Infrastructure | Retailers move away from purely visual virtual storefronts toward semantic cataloging, embedding behavioral and contextual metadata directly into product data. OpenUSD-based technical blueprints allow agents to interact with physically accurate digital twins of products. | Buying agents can efficiently “read” inventory, evaluate spatial compatibility, and compare options programmatically. This shift introduces Agent Engine Optimization (AEO), where brands structure data so agents can parse, rank and transact, similar to how SEO once optimized for search engines. |
| Agentic-Augmented Supply Chain | The Industrial Metaverse connects logistics providers through interoperable digital twins that represent each step of the supply chain in shared, three-dimensional space. | AI agents gain real-time visibility into inventory, transportation and fulfillment environments, allowing them to autonomously negotiate solutions, reroute shipments, or secure expedited transport in response to disruptions, without human intervention. |
Summary: Will the Metaverse Survive?
Despite many declaring it beyond hope, the metaverse is finding its true utility as it rapidly evolves from a consumer entertainment concept into the operating system for the agent-to-agent machine economy. This transformation represents a new chapter in agentic customer experience, where autonomous systems handle complex transactions on behalf of consumers.
While the "metaverse" moniker may be shed in favor of "Spatial Computing" or the "Agentic Web," the fundamental infrastructure of an essentially mirrored digital world will remain. As agentic AI for marketing continues to mature, the metaverse's ultimate success depends not on how much time humans spend inside it, but on how effectively we can design it for the machines that will run many aspects of our world. The future of the metaverse is no longer a place humans go to escape, but the invisible infrastructure that will make our new AI-augmented reality work.
Learn how you can join our contributor community.
