From Content Machines to Commerce Engines
For the past two years, the AI conversation has been dominated by a single use case: content generation. Write this email. Summarise that report. Draft a blog post. Useful, certainly. But it was always the appetiser.
The main course is arriving now, and it tastes like money.
AI is moving from writing to spending. The same systems that learned to generate text are now learning to browse products, compare prices, make reservations, and complete purchases. This is not a theoretical roadmap buried in a research paper. It is happening in production, right now, across the platforms that hundreds of millions of people use daily.
And the companies building the infrastructure for this shift — primarily OpenAI and Google — are locked in a protocol war that will determine who controls the pipes of AI-powered commerce for the next decade.
The Transaction Layer Nobody Is Talking About
Consider what ChatGPT has become. It is no longer just a chatbot. With browsing, plugins, and an expanding set of integrations, it functions as a general-purpose agent that can take action on your behalf. Ask it to find you a flight. Ask it to compare insurance quotes. Ask it to order dinner.
The numbers back this up. ChatGPT now commands roughly 64.5% of web traffic share among AI platforms. That is not just a lead. That is market dominance at a scale that makes it a primary channel for discovery and, increasingly, for transactions.
Google, meanwhile, is integrating Gemini across Search, Shopping, Maps, and Android in ways that make the AI layer feel native rather than bolted on. When you search for a product on Google and Gemini surfaces a comparison with prices and a “Buy” button, the line between search and commerce evaporates entirely.
Both companies understand the same thing: the platform that becomes the default interface for AI-assisted purchasing will capture an enormous share of global commerce. And both are building proprietary protocols to make that happen.
The Protocol War: ACP vs. UCP
This is where it gets technical, but it matters — because protocol wars have a way of shaping entire industries for decades. Think HTTP. Think TCP/IP. Think what Visa and Mastercard did to payment rails.
OpenAI is building what is called ACP — the Agent Commerce Protocol. The concept is straightforward: a standardised way for AI agents to interact with merchants, browse catalogues, initiate transactions, and complete purchases on behalf of users. If ACP becomes the standard, every business that wants to sell through AI will need to integrate with OpenAI’s infrastructure.
Google is countering with UCP — the Universal Commerce Protocol. Same goal, different architecture, different incentives. Google’s version leans on its existing dominance in shopping ads, merchant feeds, and payment infrastructure. UCP is designed to work across Google’s ecosystem — Search, Shopping, Maps, YouTube — giving it a distribution advantage that OpenAI cannot match.
Sitting beneath both is MCP — the Model Context Protocol — which is emerging as a kind of baseline interoperability layer. Think of MCP as the foundation that allows AI models to understand context, maintain state, and interact with external systems. Both ACP and UCP build on top of MCP, but they diverge sharply in how they handle the commercial transaction layer.
The strategic implications are significant. If ACP wins, OpenAI becomes the toll booth for AI commerce. Every transaction that flows through a ChatGPT-initiated purchase generates data and potentially revenue for OpenAI. If UCP wins, Google extends its advertising and commerce empire into the AI era seamlessly. If neither wins cleanly — which is the most likely outcome — businesses will need to support both, adding complexity and cost.
What This Means for Businesses
If you run a company that sells products or services online, this is not an abstract industry analysis. It is a preview of your next three years.
Your product needs to be discoverable by AI agents, not just humans. Today, you optimise product pages for Google’s crawlers and human shoppers. Tomorrow, you will need to optimise for AI agents that parse your catalogue, evaluate your pricing, and decide whether to recommend you — all without a human ever visiting your website.
Structured data becomes existential. AI agents rely on clean, structured product data to make purchasing decisions. If your product feeds are messy, your pricing is buried in JavaScript, or your inventory data is stale, AI agents will skip you in favour of competitors whose data is machine-readable.
You will need a protocol strategy. Just as businesses had to decide how to approach iOS vs. Android, or Amazon vs. direct-to-consumer, you will need to decide how to engage with ACP, UCP, or both. Early movers will have an advantage in shaping how their products appear in AI-mediated commerce flows.
The customer journey is collapsing. The traditional funnel — awareness, consideration, purchase — assumed that humans moved through stages over time. AI agents compress that entire journey into seconds. A user says “Find me running shoes under $150 with good arch support” and the agent goes from discovery to recommendation to checkout in a single interaction. If your brand is not in the consideration set that the AI generates, you never had a chance.
The Bigger Picture
Step back and look at the arc. In 2023, AI wrote emails. In 2024, AI browsed the web. In 2025, AI started making purchases. By 2027, a meaningful share of all online commerce will be AI-mediated — initiated, researched, and completed by agents acting on behalf of humans.
This is not a feature update. It is a channel shift. The last time something this fundamental happened to commerce was the rise of mobile shopping in the early 2010s. The businesses that adapted early — optimising for mobile, building apps, rethinking the checkout flow — captured disproportionate value. The ones that treated mobile as a nice-to-have spent years catching up.
The AI commerce shift will be faster and more disruptive, because the technology is moving faster and the incumbents are bigger. OpenAI and Google are not startups experimenting in a corner. They are trillion-dollar platforms rebuilding the infrastructure of how money changes hands.
The practical takeaway is this: do not wait for the protocol war to have a winner before you start preparing. Invest in structured product data now. Start testing how your brand appears in AI-generated shopping recommendations. Build relationships with the teams at OpenAI and Google who are shaping commerce integrations. And pay close attention to how MCP, ACP, and UCP evolve — because the protocols that win will define the rules of commerce for the next era.
The companies that understand this shift early will not just survive it. They will be the ones the AI agents recommend first.
