Recently, two of the biggest tech companies in the world launched personal AI assistants that can do your shopping for you.

Meta launched Muse in the US on 8 September. OpenAI followed on 29 September with something it calls dots. Both do roughly the same job. You tell them what you want, and they go off and do it: research products, compare prices, book a table, sort a trip, and in Meta’s case, check out and pay once you’ve approved it.

If you run marketing for a brand, I suspect you’ve had at least one conversation this month that started with “should we be worried about this?”

The short answer is no. The longer answer is that it deserves your attention now, while it’s still early, and that’s what this post is about.

What’s actually new

We’ve had AI chatbots for a few years. You ask a question, you get an answer. What’s changed is that these new assistants take action. They don’t just tell you which running shoes are good. They find your size, check who has them in stock, compare delivery times and get them ready to buy.

The industry calls this “agentic commerce”. You don’t need to remember the phrase. The idea is simple: more and more buying decisions will start, and some will finish, inside an AI assistant rather than on your website.

The UK reality check

Here’s the reassuring bit. Neither Muse nor dots is available to UK consumers yet, and there’s no confirmed date. UK regulators are paying close attention, and UK shoppers are cautious.

The numbers back that up. Only 19% of UK adults say they trust an AI assistant to make everyday purchasing decisions, according to ACI Worldwide and YouGov research in June. Six in ten said they’d stop using one after a single mistake.

Actual buying through AI is still small too. Eurostar has said AI tools accounted for around 1% of its bookings this summer. John Lewis says searches from AI assistants have grown from 0.3% to 2.5% in a year.

So this isn’t a cliff edge. But look at the direction of travel. That John Lewis number grew roughly eightfold in twelve months. And Adobe’s UK data shows shoppers who arrive from AI tools now convert around 20% better than other visitors. They arrive knowing what they want.

That’s why I’d rather you think about this now, calmly, than in three years when it’s urgent.

What changes for brands

Over the next five years, I think there are four shifts that matter.

Your shop window gets read by machines first.

When someone asks an assistant for “a waterproof jacket under £150 that arrives by Friday”, the assistant reads your product information before any human sees your brand. Prices, stock levels, sizes, delivery times, returns, reviews. If that information is patchy, out of date or hard to read, you won’t make the shortlist. This is the least glamorous part of marketing, and it’s about to become one of the most important.

Brand matters more, not less.

There’s a real risk here. Research from Checkout.com found that 57% of shoppers would let an assistant switch brands to get better value. If someone says “find me a jacket”, the assistant will optimise for price and availability. If they say “find me a Barbour”, you’ve already won. The job of brand building is to be the name people say out loud when they give the instruction. Brands that let themselves become interchangeable will find assistants treating them that way.

Search is changing shape.

People are already asking AI tools the questions they used to type into Google. Advertising is following them. ChatGPT now carries ads in the UK, and Google is testing new ad formats inside its own AI answers. None of this replaces search overnight, but where people discover brands is spreading out, and your plan needs to follow them.

Measurement gets harder.

If an assistant recommends you and the customer buys later, directly or in store, most reporting won’t give that recommendation any credit. You could be winning inside AI answers and not see it, or losing and not know why. That’s why we push clients to measure what’s genuinely adding profit, rather than what the platforms say they drove.

What we’d do now

You don’t need a big programme. You need to get the basics right before the UK market catches up with the US.

Check how you show up.

Ask ChatGPT, Gemini and Copilot the questions your customers ask. Are you mentioned? Are competitors? Is the information right?

Fix your product information.

Accurate prices, stock, delivery and returns details in your product feeds and on your site. This helps you in Google today and in AI assistants tomorrow.

Test, carefully. Try ChatGPT ads or Google’s AI formats with a small budget and a proper control group, so you know whether they add anything or just take credit for sales you’d have made anyway.

Keep investing in brand.

It’s your best defence against being reduced to a price.

Get your governance in order.

UK consumer law already says businesses are responsible for what their AI does. If you’re using AI anywhere customers can see it, know who owns it.

Where Open Partners stands

We’ve been working on this for a while. We start every piece of work with the same three questions: the right audience, the right message and the right moment. AI assistants don’t change those questions. They add a new audience (the assistant itself), a new place where the message needs to land, and a new moment that often happens without your customer watching.

Our Data, Creative and Media teams work together on exactly this, and we use AI across the agency every day, with people making the calls. We’re already helping clients tidy product data, track how they appear in AI answers, test the new ad formats and measure what actually drives profit.

Muse and dots will arrive in the UK at some point. When they do, the brands that did the dull work early will be the ones the assistants pick.

Find out how your brand shows up in AI answers:

If you’d like to talk about what this means for your business, get in touch. We’re happy to run a quick check on how your brand shows up in AI answers today. Just get in touch below.