All I Want for Christmas is AI: The tech giants would love to handle your holiday shopping

Reports from the North Pole suggest that Santa has replaced every elf in the workshop with AI this year. He also seems to have struck deals with all the major American tech giants. At least that’s the impression you get when almost all of them roll out new shopping-focused features right before the happiest season of the year for any e-commercer: the Christmas rush. The real question is whether this feels more like a present for the users, or for the tech giants themselves

The 2025 edition of Black Friday could just as well have been renamed Black Test Day - at least in the US.

Brands and retailers finally got their first real signal of how consumers are responding to Agentic Commerce and whether they’re actually interested in using AI-powered shopping features like semantic search, visual search, instant checkout and AI agents.

This followed a flood of new, consumer-facing releases from just about every AI provider out there. Google introduced agentic checkout and an AI agent that can call stores to check if a specific size is in stock. Perplexity rolled out personalized AI memory to help the model better predict what products you’re looking for. Amazon launched an AI price tracker that automatically purchases a product when it hits a certain price point. And most recently, OpenAI dropped “shopping research,” designed to eliminate the frantic ritual of juggling forty browser tabs every time you need a new frying pan.

All signs point to an official kickoff to a fundamentally reshaped customer journey. One where consumers don’t hunt for products, but where AI delivers the right options at the right price. It’s a shift we’ve highlighted repeatedly on this channel through our work with semantic search, AI agents and the universal cart - the building blocks of what we call Agentic Commerce.

The timing is hardly a coincidence

Let me rephrase… the timing is obviously no coincidence. The holiday season is the high season of all high seasons in e-commerce, so it makes perfect (financial) sense that AI providers, along with major retailers like Target, Amazon and Walmart, are eager to “help” Americans buy their gifts on their platforms.

And Americans seem more than happy to accept the AI-powered holiday assistance. Adobe Analytics expects AI-assisted shopping to jump by 520% in the US this Christmas. Shopify isn’t just predicting, they’ve looked back at the year and report that traffic to webshops via AI tools has increased sevenfold since January, and AI-driven orders have multiplied by eleven.

All of this suggests that Agentic Commerce is here to stay - and that 2026 is shaping up to be the year it really accelerates.

But even though this is an exciting shift for anyone working in e-commerce, a set of big, unanswered questions hides behind AI’s overly friendly exterior. Time to unwrap those.

What do we actually know about how it works?

Even though all these new AI features are live, and both American companies and consumers seem thrilled, there’s still surprisingly little transparency around how the systems operate behind the curtain.

How do AI assistants make their choices? How are your products evaluated in ChatGPT versus Gemini? And even more importantly: how do they get picked to appear in the results in the first place? Is the ranking fair, or are some retailers being quietly sidelined?

This is where the story shifts from Christmas magic to something a little more Matrix. As AI assistants become the new gatekeepers in the customer journey, the balance of power moves with them. And we still don’t know the mechanisms behind their recommendations. We also don’t know whether brands or retailers will eventually have to pay to secure a premium position in the AI’s suggestions. What we do know is that OpenAI is very likely to introduce advertising in ChatGPT in 2026 - and it would be no surprise if the other platforms follow suit. That thought puts a few scratches in the otherwise shiny, user-friendly, personalised experience Agentic Commerce promises.

And we probably shouldn’t expect AI platforms to be entirely unbiased in their product recommendations. For instance, Amazon recently blocked Perplexity’s Comet agent from its platform, citing violations of its terms of service - a move Perplexity publicly labelled as bullying. In practice, this means you, as a Perplexity user, are unlikely to see Amazon products recommended at all, and your Comet agent certainly won’t be able to buy them for you. Chances are this is just one example of many similar dynamics at play as tech giants jockey for consumer attention and market share.

Then there’s the web of partnerships tying the major players together: Shopify with OpenAI, Amazon with Anthropic, Meta with Microsoft. So the big question becomes whether ChatGPT, for example, will end up favouring products from Shopify-based stores. And that’s not a trivial question. It’s the difference between an open market and a semi-closed one, where money and connections shape the landscape. Not exactly a plot twist. Such a market would make it harder for smaller retailers with modest budgets to compete. But for now, much of this remains speculation.

A Christmas tale with an open ending

The conclusion is this: while it’s clear to most that Agentic Commerce is reshaping the ecommerce landscape as we know it, we’re still navigating territory where much remains uncharted.

For Danish ecommerce companies, it can be frustrating not knowing exactly which direction things will take. But there are steps you can already take to strengthen your chances of success in an AI-driven customer journey:

• Your product data needs to be structured and consistent across platforms. If you sell both through your own site and through wholesalers, it’s crucial that attributes and specifications match everywhere. Otherwise, the AI may misinterpret your product — or fail to recommend it entirely because the data isn’t aligned and up-to-date.
• The distinction between owned, earned and paid media becomes irrelevant for AI. When the AI decides which sources to trust for product recommendations, it may lean more on other consumers than on your expensive campaigns. That means owned and earned media may become even more important moving forward.
• SEO keywords will feel like relics from another era. Users will articulate their needs in concrete situations: “I need water-resistant earbuds with solid noise cancellation that stay in place when I run.” Product descriptions will need to reflect that — focusing on the problems the product solves, not just keyword-stuffed listings.

In short: AI may take over the search and recommendation work, but you’re still responsible for giving it something worth recommending.

Beyond that, consider whether it makes sense to incorporate more customer-facing AI into your solutions, for example by making your search field semantic or implementing a chatbot that can genuinely assist during the inspiration and consideration phases, not just with customer service. There are plenty of opportunities to improve the customer experience with AI while simultaneously working to be chosen by the big AI tools. Much like you’ve spent years optimising for Google while optimising your own platforms.

So even if Santa has loaded his sleigh with AI and taken it for a test flight across the US, the path ahead is still wide open. We don’t yet know whether AI assistants will remain personalised helpers or morph into algorithmic sales agents with hidden motives. We don’t know which platforms or technologies will dominate - though the smart money is on the tech giants, once again.

The real question for ecommerce teams is no longer whether Agentic Commerce will change how people shop. That’s a given. The question is whether you’ll adapt your strategy to a world where AI makes more and more of the decisions in the buying journey - or insist on stuffing your product descriptions with keywords that both users and AI ignore in favour of needs, situations and context.

Luckily, that part is up to you.