If you run a UK insurance brokerage, you’ve probably sat through at least one AI pitch in the past year. Some of the tools on offer are genuinely useful. Others are automation products with AI branding added for 2024. The gap between the two is not always obvious when you are being sold something.
What is AI actually doing in UK brokerages right now?
The Bank of England and FCA surveyed UK financial services firms in 2024 and found 75% already use AI, with another 10% planning to within three years. For insurance brokers, active adoption clusters around four areas: internal workflow automation, client service support, claims triage, and fraud screening. A third of implementations run through third-party providers.
Applied Systems, which supplies management systems to UK brokerages, reports that virtual assistants are already used for data auto-population in renewal and new-business workflows, and for capturing first notification of loss. Staff time shifts from form-filling to client-facing work. For a smaller brokerage that competes on service, that shift is where the value sits.
On the client-facing side, chatbots and AI voice bots handle initial enquiries and route calls. The FCA itself uses a voice bot to direct consumer calls to the correct body, which illustrates how regulators view this kind of use: acceptable when properly governed.
Document handling sits alongside operational automation. The UK government’s 2024 AI Sector Study notes that generative AI tools for document drafting, search, and summarisation are already widely used across UK professional and financial services firms. If you are on Microsoft 365, Copilot gives you access to these capabilities through your existing subscription.
Why does this matter for your firm?
Internal process optimisation is the top AI use area in UK financial services, cited by 41% of firms in the Bank of England and FCA survey. For a small brokerage, that is where the value lands first. Staff time recaptured from data entry and renewal admin goes back to client relationships, which is where a smaller broker wins.
The practical maths is straightforward. If account handlers spend two hours a day on admin that AI tools can handle, recapturing that time is equivalent to adding capacity without the overhead. The gains compound as platforms add more AI features to existing subscriptions.
For brokerages of five to fifty people, the most practical entry point is through tools already in your technology stack. Your broking management system, your CRM, and Microsoft 365 are all adding AI capabilities incrementally. You will encounter these features through standard updates rather than a separate procurement exercise.
Where will you actually encounter these tools?
For owner-managed brokers, AI arrives most often as a feature in software you already pay for. Your broking management system, CRM, and Microsoft 365 subscription are the most likely points of first contact. Vendors add AI features through updates rather than standalone products, so your encounter with it tends to be gradual rather than a single deployment decision.
Four practical categories cover the main use cases in a brokerage context.
Document drafting and search covers tools like Microsoft Copilot, which can draft renewal emails from prior correspondence, summarise policy documents, and search across client files. This is available through a standard Microsoft 365 business subscription, with no separate vendor relationship required.
CRM and marketing automation includes AI-driven next-best-action suggestions based on renewal dates and client history, propensity-to-buy modelling, and automated follow-up sequences. These tend to arrive as feature releases within your existing platform rather than as a separate purchase.
Claims capture via an out-of-hours chatbot is a contained and low-risk starting point. The bot collects first notification of loss details via web or WhatsApp, logs them into your broking system, and staff pick up from there. The human oversight point is clear.
Fraud and compliance screening, including sanctions checks and anomaly detection, is an area where AI adds speed to processes that already exist. For smaller brokers, using built-in analytics from your insurer or platform is more practical than building your own.
When should you move forward, and when should you hold back?
Start with internal automation, where the risk is low and the gains are immediate. The area to approach with more care is anything that touches automated decisions about individual clients. Only 2% of AI use cases in UK financial services are fully autonomous; the other 98% keep a human in the loop wherever a client could be materially affected.
Three areas to approach with care, based on the regulatory and evidence picture.
Feeding client personal data into consumer-grade AI tools is a genuine compliance risk. UK GDPR applies to any AI processing of client data, and you remain responsible for what third-party providers do with data you share. Consumer tools often use inputs to train future models. Use business-grade tools with proper data processing agreements in place.
Pricing or underwriting AI without proper governance creates regulatory exposure. The EU AI Act classifies many insurance pricing and underwriting systems as high-risk, and those obligations may reach you if you work with EU clients or use EU-supplied tools. The FCA’s Consumer Duty creates equivalent domestic pressure: opaque or biased AI-influenced pricing is a fair treatment problem.
Assuming AI output is accurate without checking it is the third area to watch. The UK government’s AI Sector Study notes that generative AI can produce inaccurate or fabricated outputs. The FCA’s outcomes-based approach means you are responsible for the effect on clients regardless of what tool produced the content.
Agentic AI, meaning systems that complete end-to-end processes autonomously, is being trialled in UK financial services. RSM’s 2026 analysis notes this as an emerging trend, but also that scaling autonomous AI agents requires redesigned governance and risk controls. For a smaller brokerage, this is worth monitoring rather than acting on now.
What do regulators expect from brokers using AI?
The FCA takes a technology-neutral approach: if you need to meet Consumer Duty doing something manually, you meet the same obligation when AI is involved. Senior managers remain accountable under SM&CR for AI activities in their area, and the FCA expects firms to test and monitor AI systems and to be able to explain how they work when asked by supervisors.
The ICO’s guidance is equally direct. UK GDPR applies in full to any AI tool that processes client personal data. You need a lawful basis, you must minimise the data you share, and you must keep it accurate. Where an AI tool makes decisions that significantly affect a client, Article 22 of UK GDPR gives that client the right to human intervention and an explanation. Meaningful human oversight at those decision points is not negotiable.
Treat AI tools as additional systems within your existing cyber controls: access management, logging, and incident response. The NCSC’s guidance on secure AI development is the relevant reference. The NCSC also notes that generative AI is increasing phishing sophistication, which matters for any firm handling client financial data.
If you broker risks for EU-resident clients or use AI tools supplied by EU providers, check whether those systems fall into the EU AI Act’s high-risk category. Your supplier carries the primary compliance obligation for the tool, but you need to understand what you are relying on.
The governance step that supervisors are increasingly looking for is simple: name the person in your firm who is accountable for AI decisions and write it down. The Bank of England and FCA survey found that board-level ownership of AI risk is now an expectation, not a recommendation.
The tools worth acting on now are those embedded in software you already pay for, applied to internal processes where the regulatory complexity is low. Claims capture, renewal admin, document drafting, and CRM automation are all available today, and the business case is clear. If you want to work through which tools fit your firm and what governance to put around them, book a conversation.
