Agentic Commerce Is Coming for D2C: What AP2 and ACP Mean for Indian Brands in 2026

Every few years the internet moves the checkout button somewhere new. It went from desktop to mobile, from your website to marketplaces, and lately from search to AI answers. The next move is stranger than any of those: the buyer clicking "buy" might not be a person at all. It might be an AI agent, purchasing on a shopper's behalf, inside rules that shopper set once and then forgot about.

That sounds like science fiction until you notice who's building the plumbing. Google and OpenAI have both shipped open standards — AP2 and ACP — for exactly this: letting an agent pay for something without breaking the fraud, authorisation and accountability systems that keep commerce safe. When the two largest AI companies quietly agree on the need for payment rails for agents, it's worth a D2C founder's attention, not because you need to act this week, but because the groundwork that makes you "agent-ready" is the same groundwork that wins AI search and runs a tighter business today. This is a strategy-level map of what's happening and what actually matters for an Indian brand.

What agentic commerce really means

Agentic commerce is the step past recommendation. AI search already answers questions and suggests products; agentic commerce is when the agent completes the transaction. A shopper tells an agent what they want and the boundaries — a budget, a timing, a condition — and the agent finds the product, weighs the options, and buys, all without the person touching a product page.

From "get the click" to "earn the agent's choice"

This rewires the core marketing question. For twenty years the job was to get a human to your checkout and reduce friction once they arrived. In an agentic world, part of the job becomes getting chosen by an agent that's comparing options on a person's behalf — an agent that reads structured data, weighs price and trust, and never sees your beautiful hero image. Your storefront's persuasion still matters for humans, but a second, machine-facing layer starts to decide whether you're even in the running.

The rails: AP2 and ACP

The reason this is suddenly credible is that the hard part — paying safely — is being standardised.

Google's AP2, the Agent Payments Protocol, was announced in September 2025 with more than 60 partners, including Mastercard, PayPal, American Express, Coinbase and Etsy. Its clever bit is the "mandate": a cryptographically-signed, tamper-proof digital contract that proves a shopper authorised an agent to make a specific purchase. That solves the three questions autonomous buying raises — did the user authorise this, does it reflect their real intent, and who's accountable if it goes wrong — which is what payment networks needed before they'd allow agents anywhere near a transaction.

OpenAI's ACP, the Agentic Commerce Protocol, is the parallel standard, built with Stripe and powering purchases inside ChatGPT. The two aren't identical in approach, which matters for how you plan.

The takeaway from that table isn't which one to pick. It's that no one has won, so betting everything on a single protocol is the wrong move. Prepare for a world with more than one.

Why the boring work is the real strategy

Here's the part that saves you from hype. You don't need to implement a protocol, and you shouldn't try to. What decides whether an agent can transact with you is far less glamorous: whether your product data is clean enough for a machine to read, trust and act on.

An agent can only buy what it can understand. If your price is stale, your stock is wrong, your specs are vague, or your feed drifts out of sync with your store, an agent has nothing reliable to choose from and will pick a competitor whose data it can trust. Your product feed stops being a technical afterthought and becomes a storefront in its own right — one that a machine reads instead of a human. This is exactly the kind of full-funnel foundation our growth marketing team treats as core infrastructure, because a clean catalogue and honest signals pay off in Shopping, in AI search, and in whatever agentic future arrives.

The India reality: early, but not optional

Be honest about timing. Real consumer adoption of agent-led buying in India will take a while, and when it comes it'll largely ride on rails people already trust, like UPI, rather than anything exotic. So this isn't a reason to panic or to chase a shiny integration. It's a reason to get your fundamentals clean now, because every one of them pays off immediately and also happens to be what an agent will need later.

What to actually do this year

The agent-ready checklist is the same list that makes you better today, which is why it's worth doing regardless of how fast agents arrive. Keep your product feed accurate and machine-readable — titles, prices, variants, specs — and in sync across your store and Merchant Center. Add thorough structured data so your catalogue is unambiguous to a machine. Make your trust signals legible with genuine reviews, clear shipping and return policies, and consistent brand information across the web. And keep inventory and fulfilment honest, because an agent buying an out-of-stock product fails a customer harder than a human doing the same. Every item there improves your AI-search visibility and your operations now, and quietly makes you buyable by an agent later.

What not to do while it's still early

The failure modes here are as predictable as the opportunity. Don't rush to bolt on a protocol integration you don't need yet — the standards are still moving, and an early bet on one rail can age badly before real adoption arrives. Don't treat agentic commerce as a reason to neglect the human experience on your site, since people will remain the overwhelming majority of buyers for a long while. Don't let vendors sell you a pricey "AI commerce" product whose value is really just the clean feed and structured data you could build yourself. And don't ignore it entirely either — the brands that dismissed AI search a year ago as hype are now scrambling to be cited, and the pattern will repeat with checkout. The balanced posture is to invest in the fundamentals that pay off today and keep you ready, while watching the protocols mature before committing to any one of them.

The takeaway

Agentic commerce moves the checkout from the human to the agent, and the rails — Google's AP2, OpenAI's ACP — are being laid now by the biggest players in AI and payments. For Indian D2C brands the smart posture isn't to implement a protocol or to dismiss the whole thing as hype; it's to recognise that being agent-ready and being excellent today are the same work. Clean, accurate, machine-readable product data, honest stock and fulfilment, and strong trust signals win you AI-search citations and Shopping performance right now, and they're precisely what an agent will need to choose and buy you when adoption tips. Get the fundamentals clean, stay protocol-agnostic, and you're ready for the next place the internet moves the buy button — even if this time nobody's finger is on it.

Sources: Google Cloud Blog — "Announcing Agent Payments Protocol (AP2)" (announced September 2025; extends Google's Agent2Agent and Model Context Protocol; uses cryptographically-signed Intent and Cart 'mandates' to establish authorisation, authenticity and accountability; 60+ partners including Mastercard, PayPal, American Express, Coinbase and Etsy; supports cards, stablecoins and bank transfers). Reporting on OpenAI's Agentic Commerce Protocol (ACP), built with Stripe, powering purchases inside ChatGPT. Strategic interpretation, the India/UPI framing and the agent-readiness checklist are Digistex4u's applied guidance.