Most brokers do not have an AI strategy even though they are using it
The following article was supplied by Moterra.
Most brokers do not have an AI strategy yet. Most broking teams already use AI anyway. Renewal summaries, market submissions, policy wording comparisons and claims updates are increasingly drafted with public AI tools, usually by people trying to keep up with the workload rather than to break any rules. The wider market is moving the same way: the Bank of England and FCA survey published on 26 October 2026 found that 86% of UK financial services firms now use AI, up from 75% in 2024.
That creates a gap between how the firm thinks it operates and how it actually operates. For a regulated business, that gap is the problem.
The compliance questions
The FCA has said it will oversee AI through the existing frameworks, and each of them applies to a broker using AI on client work:
- Consumer Duty. For retail business, the firm must show that AI-assisted outcomes are fair and the advice suitable.
- SM&CR. Responsibility for how the firm uses AI sits with a named senior manager, whether or not anyone formally approved the tools in use.
- Operational resilience. Once AI supports renewals, claims or client communications, it is part of an important business service, so the firm must know where it runs and which third party sits behind it.
- UK GDPR. A prompt containing client details is personal data, with the same duties as any other processing.
Brokers with EU operations, or placing business with EU insurers, meet the same questions under DORA. Public AI tools fail all of them: client information leaves the firm, nobody can say where it is processed, and there is no record of what was asked or who checked the answer.
Bringing AI inside the firm
Banning AI rarely works. Staff move to personal phones, and the firm loses even the little visibility it had. The better route is an approved tool that runs inside the firm’s own cloud environment, in a UK region, so client data never leaves the business.
That is what Moterra does for brokers and MGAs. We deploy Claude, Anthropic’s AI model, inside the firm’s own AWS or Azure account, encrypted with the firm’s own keys and never used to train a model. Because the environment belongs to the firm, AI sits inside the cloud arrangements its resilience and third-party oversight already cover, rather than becoming a new supplier to explain. The controls then follow naturally:
- Access by role. Connectors to email, document storage and broking systems inherit existing permissions, so account handlers see their own clients and claims staff share only what a claim requires.
- A record of use. Every request and output is logged, so the firm can show who used AI, on what and when.
- A human sign-off. AI can draft a renewal summary, but a named person approves anything that reaches the client.
“Brokers are not short of AI, they are short of AI they can stand behind,” said Vytenis Sakalas, CEO and co-founder of Moterra. “When a capacity provider or the FCA asks how AI is used on client work, the firm should be able to answer in one sentence: it runs inside our own cloud, under our own controls, and a named person signs off what reaches the client.”
Keeping control of cost
The second surprise is usually the bill. Pilots start with a few enthusiastic users. When the whole office wants access, licences priced per seat, charging the same whether someone uses AI daily or once a month, quickly outweigh the benefit.
Moterra gives the firm one admin console for access and spend. There are no per-seat Claude licences: the firm pays for actual usage, sets spending limits by team, and sees adoption, cost and time saved on a single dashboard. Finance gets a predictable figure, and the board gets a clear answer to the question it eventually asks: who approved this spend, and what did it buy?
Where to start
- Find out what is already happening. Ask teams which tools they use and for what. The honest answer is more useful than the official one.
- Decide where an approved tool should run and who should be able to see the data it handles.
- Agree access groups and a spending limit before rollout, not after the first invoice.
A standard Moterra deployment is typically live in one to two weeks, so a firm can move from informal use to controlled use well before the next renewal peak. Done in that order, AI stops being a compliance risk nobody has measured and becomes a tool the whole office can use with confidence.
About Moterra
Moterra is a private AI deployment company and a Select Services Partner in the Claude Partner Network. We help businesses deploy Claude inside their own AWS or Azure environment, connect it to the tools they already use, and keep it secure and governed, so everyday processes get faster.
Media contact
Amanda Uzialaite, CMO, Moterra, [email protected]
