AI shopping agents are now completing real purchases. In 2026 they operate inside ChatGPT, Google Gemini, Microsoft Copilot, and Perplexity, comparing offers and buying on a customer's behalf. Retailer traffic from AI sources has surged around 1,200% while traditional search traffic fell about 10% year on year.1 Google launched a Universal Commerce Protocol at NRF in January 2026,2 and OpenAI's Agentic Commerce Protocol, built with Stripe, already connects Instacart, DoorDash, Shopify, and Etsy.2 Morgan Stanley expects nearly half of online shoppers to use AI agents by 2030, accounting for roughly a quarter of their spending.3 AI shopping agents are already buying on customers' behalf. Decide whether your product, pricing, and availability data can be found and acted on by the algorithms doing the shopping.
Why It Matters
For commerce and operations leaders, this changes who the customer actually is. When an agent does the buying, it does not respond to storefront design, promotional banners, or brand campaigns. It reads structured data through an API and decides on attributes such as price, materials, durability, and availability. Commerce shifts from managing channels for human shoppers to making products legible and competitive to machines that buy.
The risk of treating this as a distant trend is being invisible to the buyer. If product data is built for a human browsing a website rather than an agent querying an API, an AI shopper will not surface you at all. Brand loyalty offers less protection than it did, because agents compare on attributes rather than familiarity.
This reaches further into the operating model than the storefront. If agents transact through open protocols, pricing has to be machine-readable and current, and availability data has to be accurate to the unit. Each of those is an operations decision, not a marketing one, and each exposes whatever gaps already exist between systems. The retailers who win the agentic channel will be the ones whose product, price, and fulfilment data are good enough to be acted on automatically.
6xD Interpretation
- Primary lens — D1: Economy 4.0. Agentic commerce is a structural shift in how value and demand form in retail — a new competitive layer, not a new channel.
- Supporting lens — D3: Digital Business Platforms. Being legible to an agent requires product, pricing, and fulfilment data exposed through APIs on a coherent platform, not scattered across systems.
- Supporting lens — D6: Digital Acceleration Tools. The open commerce protocols (Google's, OpenAI's) are themselves acceleration mechanisms that define how fast a retailer can plug into the agentic channel.
6xD Insights interpretation: The question for retail is blunt — can the algorithm doing the buying actually find you? Brands it cannot read are brands shoppers will never be offered.
Executive Implications
| Decision area | Executive question | Required output |
|---|---|---|
| Data readability | Can an AI agent read and act on your product, pricing, and availability data through an API? | Machine-readable, current product and pricing data |
| Protocol participation | Is the organisation connected to the open commerce protocols agents already use? | A defined position on Google's and OpenAI's commerce protocols |
| Data ownership | Who owns fixing the gaps between systems that block agent-readable data? | A named owner for agent-readable commerce infrastructure |
| Competitive exposure | Would your products be surfaced to an AI shopper today? | An audited answer, not an assumption |
Recommended Actions
- Audit agent readability now. Check whether your product, pricing, and availability data can be read and acted on by an AI agent through an API.
- Fix the gaps before competitors do. Prioritise the data gaps that make you invisible to agents, ahead of the retailers already positioning.
- Take a position on the open protocols. Decide how the organisation connects to Google's Universal Commerce Protocol and OpenAI's Agentic Commerce Protocol.
- Own it as infrastructure, not marketing. Assign agent-readable commerce to an operations owner, since pricing and availability accuracy are operations decisions.
Sources
- 012026 agentic-commerce reporting (live AI purchase agents; approximately 1,200% AI-source traffic rise). ↩
- 02Google Universal Commerce Protocol (NRF, January 2026); OpenAI Agentic Commerce Protocol with Stripe. ↩
- 03Morgan Stanley AI-shopper adoption forecast (approximately 50% of online shoppers by 2030). ↩
