The FCA just published a document that describes a future where your client’s AI agent manages their mortgage, their savings, their insurance and their investments without them lifting a finger. It is called the Emerging Technology Horizon Scan 2026, it landed in June, and it is worth reading carefully. Not because it should worry you. Because it tells you exactly where your value as a broker is heading.
The regulator’s central concept is what it calls the “proxy economy”. In this model, consumers increasingly delegate their financial management to AI agents that search, compare, negotiate and transact on their behalf. The consumer sets the boundaries. The AI does the work. Everyday financial management is conducted through machine-to-machine interactions that humans only occasionally review.
If that sounds abstract, consider what it looks like in practice. The FCA describes an AI agent that manages household finances within agreed boundaries, moving money to higher-rate savings, flagging uncompetitive insurance renewals, even switching current accounts. All autonomously. All at machine speed.
The instinctive reaction for most brokers will be to ask whether this makes them redundant. The honest answer, drawn directly from the FCA’s own framework, is quite the opposite.
Three stages, and why the third one needs you
The FCA describes the proxy economy emerging in three stages.
In the first stage, assistive mode, AI explains products, compares options, pre-fills forms and flags risks. The consumer still makes every decision. This is roughly where we are now with the better sourcing tools and AI-assisted fact finds.
In the second stage, advisory mode, the AI recommends specific actions — switching suppliers, refinancing, changing savings behaviour — and the consumer typically clicks through the suggestions. Think of it as a very good comparison site that also acts on what it finds.
In the third stage, “do it for me” mode, the AI is authorised to act within dynamic constraints. It autonomously negotiates prices, optimises bills, reallocates investments, disputes charges. The consumer gets periodic summaries. The AI runs the show.
Here is the critical detail. Even in that third stage, the FCA is clear that consumer attention is not eliminated. It is “relegated to escalation cases: moments where proxies are uncertain, where stakes are high or where regulatory requirements demand explicit consent.”
Read that sentence again, because it describes the mortgage market almost perfectly.
Why mortgages are escalation territory
A mortgage is not a savings account that can be optimised in the background. It is the largest financial commitment most people ever make, tied to a property they live in, a family they are raising, a future they are planning. It involves regulatory requirements for explicit consent. It involves high stakes by any reasonable definition. And it frequently involves uncertainty — complex income, unusual property types, changing personal circumstances, later life considerations — that an AI agent cannot resolve alone.
The FCA’s framework essentially says: routine financial management will be automated, but the big, complex, consequential decisions will still need a human. Mortgages are exactly that.
This is where the broker’s value is heading. Not towards sourcing the cheapest two-year fix. An AI can do that faster and cheaper than any human. The value moves towards the conversation the AI cannot have: the one about whether the client should fix for five years because they are planning to start a family, or whether they should overpay now because they can afford to and want to be mortgage-free before the children start university, or whether equity release is the right answer for their parents.
The brokers who will thrive in a proxy economy are the ones who become the human the AI calls when it reaches the limit of what it can do alone.
What “human in the loop” actually means
The phrase “human in the loop” gets used a lot, usually in a vague, reassuring way. The FCA’s Horizon Scan gives it a sharper meaning.
In a proxy economy, a well-positioned broker is not competing with the AI. They are part of the system. The AI handles the research, the rate comparison, the document gathering, the initial affordability assessment. The broker handles the judgement: is this the right product for this client’s actual situation? Are there risks the data does not capture? Is the client’s stated preference aligned with their real needs?
The FCA describes this as the “human + AI” model, where an AI assistant acts as an extension of the individual. A digital twin that understands the consumer’s financial position but escalates to a human when the decision requires empathy, nuance or regulatory sign-off.
For brokers, this means the admin shrinks and the advice expands. The hours currently spent chasing documents, keying data into portals and waiting for lender responses get absorbed by the technology. The hours spent understanding a client’s life, asking the questions they had not thought to ask, and making recommendations that reflect genuine understanding of their circumstances — those hours become the entire job.
That is not a diminished role. It is a more valuable one.
The risk of standing still
The FCA’s Horizon Scan also describes a less comfortable scenario. If Big Tech firms use AI to capture key parts of the financial value chain, and consumers delegate their financial management to those platforms, then brokers without a strong, integrated proposition could find their clients being switched away before they even know it happened. The proxy does not call to discuss it. It just acts.
The defence against that is not to resist the technology. It is to make your advice so embedded in the client’s financial life that the AI agent has no reason to look elsewhere. A broker who understands a client’s mortgage, their protection, their long-term wealth plan, their family circumstances — that broker has built something an algorithm cannot easily replace.
The FCA notes that trusted proxies could help consumers manage complex financial affairs that are increasingly fragmented across platforms and products. The broker who brings those fragments together, who provides the coherent view that the AI agent draws on, is not being replaced. They are becoming more important.
Where this leaves you
The proxy economy is coming. The FCA would not have published a 26-page horizon scan about it if they thought otherwise. The Mills Review, launched in January 2026, is already examining how AI is being used across retail financial services. The direction is set.
But the direction favours good brokers, not replaces them. The routine work will be automated. The complex, consequential, deeply human work will not. The question is whether you are positioning yourself for the work that stays or clinging to the work that goes.
The brokers who read the FCA’s Horizon Scan and see a threat are looking at the wrong part of the picture. The ones who see it and recognise their own value reflected back at them — in the escalation cases, in the high-stakes decisions, in the moments where the AI reaches its limit and needs a human it can trust — are the ones who understand where this is heading.
The AI will handle the process. You handle the person. That has always been where the real value sits. The proxy economy just makes it explicit.
Research notes
FCA, “Emerging Technology Horizon Scan 2026,” June 2026. https://www.fca.org.uk/publication/corporate/emerging-technology-horizon-scan-2026.pdf
FCA, “The FCA’s long term review into AI and retail financial services: designing for the unknown.” https://www.fca.org.uk/news/speeches/fca-long-term-review-ai-retail-financial-services-designing-unknown
FCA, “Review into the long-term impact of AI on retail financial services (The Mills Review),” January 2026. https://www.fca.org.uk/publications/calls-input/review-long-term-impact-ai-retail-financial-services-mills-review
FCA, “AI and the FCA: our approach.” https://www.fca.org.uk/firms/innovation/ai-approach