Key Takeaways
- commercetools introduced Autonomous Commerce, a new category in which AI runs operations such as pricing, inventory, and promotions in real time
- The newly unveiled Sphere platform powers over $100 billion in annualized GMV at sub-60ms latency, governing each AI agent with identity, scope, and human approval
- For merchants, this splits the investment decision in two: responding to customer-facing agents versus entrusting AI with internal operations
commercetools Declares Autonomous Commerce a New Category
commercetools declares 'Autonomous Commerce' the next major era in commerce in which AI systems make and execute operational decisions in real-time.
www.prnewswire.comA pioneer of headless commerce, commercetools has framed the AI era as a distinct new category it calls Autonomous Commerce. The term describes a stage in which AI systems make operational decisions, on matters such as pricing and inventory, in real time and carry them out. The company positions this as the next major era that commerce is entering.
What stands out is that commercetools refuses to call this a mere feature. Founder Dirk Hoerig put it plainly: "Autonomous Commerce is not a feature. It is the next category of commerce." At its core sits the idea that AI senses signals, determines the right action, and carries it out, all within defined rules and boundaries.
CEO Doug McNary measured the shift against history, noting that commerce is in its biggest transition since moving from monolithic to headless systems. The implication is that infrastructure now has to be rebuilt on the assumption that both humans and AI agents are operating it.
What Separates Agentic from Autonomous
The term worth pinning down here is "agentic commerce," which is easily confused with this announcement. The two are mixed up often, yet what each one sets out to change is entirely different.
Agentic commerce describes a world where AI shops on behalf of the customer. A consumer briefs an agent on preferences, budget, and constraints, and the AI handles the search, comparison, and purchase. Orders that complete inside consumer AI like ChatGPT or Gemini fall into this category. The lead actor is the customer's agent, and the enterprise is in the position of responding to it.
Autonomous commerce, by contrast, is a world where AI runs the enterprise's operations itself. Pricing, inventory, promotions, fulfillment routing, and other core functions are decided and acted on by AI in real time. Where agentic commerce changes where customers place orders, autonomous commerce changes how the company delivers.
| Dimension | Agentic Commerce | Autonomous Commerce |
|---|---|---|
| Actor | An AI agent acting on behalf of the customer | AI systems running the enterprise itself |
| Scope | Search, comparison, and the buying experience | Pricing, inventory, promotions, and fulfillment operations |
| What it changes | Where customers place orders | How companies deliver |
| Human role | Shoppers set preferences and limits, keep final say | AI executes within rules, escalating only on exceptions |
| Primary host | Consumer AI like ChatGPT or Gemini | An AI-native foundation inside the enterprise, such as Sphere |
This distinction ties directly to the reality of investment decisions. The industry has warned that advancing customer-facing agent support without an autonomous execution layer underneath delivers efficiency in narrow areas rather than operational transformation. Confuse the two, and the throughput gains you expected may fail to materialize well after go-live. commercetools likely named a fresh category precisely to draw that line clearly.
The AI-Native Foundation and Governance Sphere Provides
Alongside the new category came the platform meant to underpin it, Sphere. Built as a headless, API-first system handling carts, orders, checkout, catalog, inventory, search, promotions, and customer management, it already supports more than $100 billion in annualized GMV, according to the company.
The emblematic figure is its speed. Average response time runs under 60 milliseconds, a number framed around AI workloads operating at machine speed rather than human interaction. Sphere is built on MACH principles, namely microservices, API-first, cloud-native, and headless, and provides a secure entry point for third-party and customer AI agents to connect.
The crux of autonomous commerce, though, is the governance model. In Sphere, each AI agent receives an identity, a defined scope, and designated points that require human approval. Enterprises define what each agent is allowed to do and how far, keeping control over permissions, policies, and operational limits. Cited examples include pricing agents that monitor and adjust prices automatically, and procurement agents that place reorders against approved supplier lists and escalate to a human only when a decision falls outside set parameters.
commercetools is also previewing MosAIc, described as the first autonomous commerce operations tool for orchestrating multiple agents. The aim is to coordinate agents across pricing, promotions, and fulfillment, starting from a desired business outcome.
From Headless to AI-Native, the Competitive Axis Shifts
This announcement deserves to be read in a context broader than one company's product strategy, because the competitive axis for commerce platforms is moving past headless readiness.
On the agentic side, Shopify has put forward the Universal Commerce Protocol (UCP) co-developed with Google, while the Agentic Commerce Protocol (ACP) from Stripe and OpenAI powers ChatGPT's Instant Checkout. Salesforce, too, is building a commerce agent layer in Agentforce. The standardization race over how to connect customer-facing AI is already well underway.
commercetools is placing its bet somewhere else. Rather than the standardization of customer touchpoints alone, it has moved early to claim, as a category, the execution foundation and governance for entrusting internal operations to AI. The strategy extends the API-first strength built through MACH architecture into a foundation that AI agents can safely drive.
That said, a cautious view persists across the industry that a fully autonomous stage, where agents complete purchases or operational decisions without human approval, remains some way off. Sphere's governance design, which builds human-approval points in explicitly, reads as an acknowledgment of that reality.
Conclusion
commercetools' framing of autonomous commerce and its launch of Sphere push the AI commerce conversation outward, from customer-facing agents toward the autonomous execution of enterprise operations. The figures of over $100 billion in annualized GMV and sub-60ms responses signal that this is tied to a live platform rather than a concept ahead of its substance.
The takeaway for merchants is clear. Responding to agentic commerce and investing in autonomous commerce should be treated as separate decisions. The former is about where customers buy, while the latter is about how much of your own pricing, inventory, and fulfillment you are willing to hand to AI, a question of operational design itself. How to grant each agent identity, scope, and human approval will become the central theme of implementation from here.
