Picture this: It's 2027. You upload your home, car, travel, and mobile phone insurance documents to your AI assistant. You add the insurance that came free with your bank account and the travel cover included with your premium credit card.
"Optimise this mess," you tell it.
Thirty seconds later, it responds: "You're paying for the same travel insurance three times. Your mobile phone excess is £150 through your standalone policy, but only £50 through your bank account. You have no home emergency cover despite owning a 100-year-old house. I can reduce your annual insurance spend by £847 while increasing your coverage. Want me to sort it?"
You click yes. The AI cancels redundant policies, purchases new ones, and sets reminders to re-optimise at renewal. You never speak to an insurance company. You never see an advert. You definitely don't care which brand you end up with.
This is one possible future. And while it might sound extreme, the technology to make it happen exists today. The question isn't whether AI will change how we buy things - it's how much it will change, and how quickly.
We're entering an era where AI agents won't just help us make decisions, they'll increasingly make decisions for us, or at least do the heavy lifting before we make a final choice.
But let's be realistic about adoption. When comparison sites launched 20 years ago, many predicted the death of brands in financial services. Why would anyone pay more for a branded insurance policy when they could see the cheapest option instantly?
Yet here we are. Comparison sites are ubiquitous, but brand still matters. People use them to research, but often end up buying from a name they recognise. Trust, reputation, and familiarity still influence decisions, even when objective data is readily available.
So why will AI be different? And will it be that different?
The answer is nuance. AI agents will operate on a spectrum:
This article focuses on the first scenario because it's the most provocative. But even if only 10% of consumers fully delegate and 50% use AI-assisted purchasing, that would still represent a fundamental shift in how brands reach customers.
Right now, payment regulations require explicit customer authorisation for each transaction. Strong Customer Authentication under PSD2 means you need to actively approve payments, which could be a barrier to AI agents making purchases on our behalf.
But regulations evolve to match behaviour, not the other way around. We'll likely see:
Pre-authorised spending frameworks where you give your AI permission to spend up to certain limits in defined categories. Think of it like a more sophisticated direct debit, you're authorising the AI to act within parameters you control.
Tiered approval systems where small, routine purchases happen automatically, but anything above a threshold gets flagged for quick confirmation. Your AI books the £80 train ticket without asking, but the £2,000 holiday package needs a thumbs up.
Detailed audit trails showing exactly why the AI made each decision. "I selected this insurer because they have a 4.8-star claims rating, process claims in 24 hours versus the industry average of 7 days, and cost £127 less annually for equivalent coverage."
The PSR, FCA and other regulators will adapt. They'll have to. The efficiency gains are too significant, and consumer demand will drive it. The question isn't if this happens, but when.
Let's go back to that insurance example, because it illustrates both the threat and the opportunity for brands.
Traditional insurance marketing is built on emotional resonance. Trust. Reliability. "You're in good hands." Clever meerkats or opera singers. Billions spent on brand building.
But when an AI is choosing your insurer - or even just presenting options - it's comparing:
Here's where it gets interesting: brand reputation feeds into several of these metrics. An insurer with strong brand recognition often has better customer service scores, more reviews to analyse, and a track record that AI can verify. The brand didn't disappear; it just got translated into objective data points.
But brands that have relied purely on marketing without backing it up with performance? They're in trouble. The AI will see through it.
This is what selective commoditisation looks like. Products that compete mainly on specifications and price become vulnerable. But products where brand signals genuine quality, reliability, or service excellence survive, albeit for different reasons than before.
Financial services face this reality across the board. Mortgages, savings accounts, and credit cards are all highly comparable once you strip away the marketing. But not all financial products are equal:
Commodity products (basic savings accounts, simple insurance policies): Highly vulnerable to AI-driven price comparison
Complex products (mortgages, pensions, wealth management): Will likely remain AI-assisted rather than AI-driven, with humans making final decisions
Relationship products (private banking, financial advice): The human relationship remains the product, though AI might help with execution
Brand isn't dying, it's evolving. Here's how smart brands will adapt:
1) Why brand still matters in an AI-assisted world
Even when AI agents are doing the research, the brand continues to provide value.
Trust signals: When an AI presents three insurance options and one is from a brand you recognise, that familiarity reduces perceived risk. The AI might rank them objectively, but you're more likely to choose the known quantity.
Quality proxies: Strong brands often correlate with better performance. AI agents will learn this. If customers consistently have better experiences with Brand A than Brand B, the AI factors that into recommendations.
Social proof: Brand strength often reflects accumulated customer satisfaction. Thousands of positive reviews and high retention rates tell the AI something valuable.
Risk reduction: For high-stakes purchases (mortgages, pensions, health insurance), brand reputation provides psychological safety even when the AI says the terms are comparable.
The difference is that brand value must be earned through performance, not just created through clever marketing.
2) Become the choice AI agents recommend
Remember how brands obsessed over Google's algorithm? SEO, structured data, site speed - all to rank higher in search results?
The same thing is about to happen with AI agents. Brands will optimise to be the choice that Claude, ChatGPT, or Apple Intelligence recommends.
This means:
The marketing question shifts from "How do we make customers love us?" to "How do we make AI agents select us and how do we make customers trust that selection?"
3) Create measurable superiority
You can't fool an AI with clever messaging. If you claim to have the fastest claims processing, the AI will verify it. If you say you're the best value, it will compare you against every competitor.
This is good news for brands that genuinely deliver. For years, inferior products have competed through superior marketing. That era is ending.
The brands that win will be those that invest in performance improvements:
These become your marketing messages, but they're not just marketing messages - they're verifiable facts that AI agents will discover anyway.
4) Own the experience layer
AI can buy an insurance policy, but it can't hold your hand when you're making a claim after your house floods.
The brands that thrive will be those that deliver exceptional experiences at the moments that matter. Because AI agents learn, if customers using Insurer A have smooth claims experiences and customers using Insurer B face bureaucratic nightmares, the AI will factor that into future decisions.
Your brand reputation becomes your performance reputation. There's nowhere to hide, but if you're genuinely good, you'll finally get credit for it.
5) Build for both humans and AI
The winning strategy isn't either/or - it's both/and.
Your brand needs to:
Think of it as marketing to two audiences simultaneously: the AI that's doing the research, and the human who's making (or approving) the final decision.
6) Accept you could be infrastructure
Some brands will realise that for certain customer segments, they're no longer consumer-facing; they're infrastructure providers that AI agents select on behalf of consumers.
This isn't necessarily bad. Infrastructure can be highly profitable, but it requires a different strategy:
Think of it like the shift from branded to white-label products. The brand on the package matters less than being the supplier that gets selected.
If you work in financial services marketing (and if you're reading a Creode blog post, you probably do), this matters more to you than most.
Financial products are particularly susceptible to AI disruption:
But the timeline and extent of disruption will vary significantly:
Quick adoption (1-3 years):
Moderate adoption (3-7 years):
Slow adoption (7+ years):
The insurance example isn't hypothetical - it's coming. But a 40-year-old remortgaging their house will probably still want to understand the options themselves, even if AI does the initial research.
Mortgages: "Find me the best mortgage for my situation, factoring in fees, early repayment charges, and likely rate changes over the fixed period." [AI presents three options, you choose]
Savings: "Spread my emergency fund across institutions to maximise FSCS protection while optimising interest rates." [AI proposes allocation, you approve]
Investments: "Build me a portfolio matching these risk parameters, minimising fees and tax implications." [AI constructs portfolio, you review before investing]
Notice the pattern? AI does the heavy lifting; humans make the final call. This is AI-assisted purchasing, and it's where most of the market lands.
For 100 years, brands have been built on emotional connections, memorable advertising, and top-of-mind awareness. The goal was to be the brand consumers thought of first when they needed your category.
In the AI age, the goal is to be the brand that AI recommends first and that consumers trust when they see it on the list.
That's a different game. It rewards performance over clever marketing. But it doesn't eliminate brand value. For brands that have relied on marketing to paper over mediocre products, this is a threat.
For brands that have been excellent but struggled to communicate it, it’s an opportunity.
The question is which one are you?
At Creode, we help financial services brands navigate digital transformation and prepare for an AI-mediated future. If you're wondering how your brand strategy needs to evolve, let's talk.
Proud to partner with
