Big Boys and Girls AI Games

Following AI developments and platforms shake ups is like reading tea leaves in a wind tunnel. No one knows quite where it’s going, but big players are making moves.

Past few weeks have been big ones.

Walmart unveiled Sparky, its conversational AI agent. Sparky’s main mission is to help help customers with real objectives, not just search queries. Think ‘find me wardrobe for a spring trip to Florence’ level of shopping projects. Walmart plans for it to interact with outside agents like ChatGPT. It’s collaborative AI at scale.

Amazon, meanwhile, is building its own AI engine. But in classic Amazon fashion, it’s walled in. Proprietary. No external agents allowed. It does not want ChatGPTs and other LLMs crawl the marketplace. The company seems to be betting that its own AI tools, like Rufus the AI shopping assistant, will be powerful enough to win without playing nice with others.

It’s a tale of two philosophies: Walmart is betting on open as a strategy to win in the AI future (which some would say IS the future). Amazon bets on dominance.

And OpenAI (ChatGPT’s parent) is quietly building its own shopping marketplace. They’re testing a “Buy with ChatGPT” feature. Fulfillment goes to the merchant. ChatGPT takes a cut. Another player steps into ecommerce with AI-native infrastructure.

I see phrases like ‘traditional search is dead’, or ‘SEO is dead’. Which sounds fit for a click-baitish post on LinkedIn. But it many ways it’s true: step of searching is now a blip of a millisecond for the computing done by AI. And knowing the ‘why’ behind the search (back to the ‘find me wardrobe for the Florence spring trip’ shopping objective) is the main interaction between a shopper and the brand.

In other words, keyword searches are becoming irrelevant (not to be confused with relevancy of copy and keywords, that’s still needed). Goal-oriented, semantic, contextual retrieval -across price, reviews, inventory, pack size, delivery timelines, is what will be happening now.

Now back to the Big Boys and Girls recent power moves.

This week also brought Amazon’s abrupt exit from Google Shopping ads. Amazon went from buying 60% of Google ads to 0% over a handful of days. DTC brands could try to capitalize on temporary vacuum (what have you seen in your Google ads this past week?). But a bigger question is, why Amazon did that. It’s not about a typical reallocation of ad spend. It;s about AI agent data access. Amazon doesn’t want Google access it’s search-rich data, after all, Google is Amazon’s big AI competitor. Maybe Amazon will negotiate Google paying a commission, or licensing type of agreement for letting Google AI agents access it’s marketplace data? TBD.

And let’s not forget about Shopify. It is often overlooked, but Shopify holds 30% market share of all e‑commerce websites in the U.S, and 1 in 6 internet users globally. Shopify quietly updated their code that basically blocks AI scrapers. In plain language: “Checkouts are for humans.” They’re slamming the door on automated browsing and agent-driven checkout flows.

As Marketplace Pulse put it well, “No one wants to be where the AI agents are shopping - everyone wants to build the AI agents that do the shopping.”

It’s platform power plays, dressed up in code and computing power.

If this sounds like an esoteric battle between trillion-dollar companies, it is. But it also touches your business more than you might think.

Because what these companies are negotiating, building, and walling off will define how discoverable your products are. And how easy they are to buy.

This isn’t a repeat of Facebook Shops or social commerce, or a big algorithm update. This is the foundational shift of who does the shopping-humans or agents.

In the end, nobody knows yet how it will shake out. But from observing and participating in the industry, here are some practical takeaways:

1. Product pages to talk like people, and answer natural questions fast. (ex: is this gluten free? yes/no).

2. Complex, too pretty infographics are out. Simple visuals, that can be grabbed and interpreted by AI correctly, are in.

3. Searches are semantic, contextual, and not keyword driven

4. AI agents are the new gatekeepers. Instead of scroll-and-click, we’ll soon see 1–2 curated product options, surfaced by AI agents based on semantic matching, past behavior, and inferred intent.

5. Platform AI wars are just beginning. Amazon vs. Walmart vs. ChatGPT vs. Shopify- each platform is angling to control not just the sale, but the buying experience. Data control = power. Which is why they’re blocking each other’s bots and building their own.

It is important for brands to watch and adapt, and test. Future of ecommerce shopping will be fragmented, meaning, many platform-centric AI agents scouting out there, working on shopping objectives and intents of consumers. But no matter how this all shakes out building a healthy business, with really good products, strong operations, having some direct channels, and as much ownership of customer data as possible will position any brand well for whatever AI-agent ecommerce world shift.

Saludos,

Irina