What changes
This is where the facilitator role starts to take shape. AI takes on the tasks that fill most of a case: reading documents, pulling figures into the fact-find, shortlisting products against lender criteria and insurer underwriting, chasing missing items and drafting the suitability letter. The adviser spends the freed time with the client, making the judgement calls and checking what AI has prepared.
The centrepiece is the evidence pack. For every recommendation, AI drafts a record of the client’s circumstances and priorities, the options considered, why the chosen mortgage or policy fits and why the others were ruled out. The adviser checks it, corrects it and signs it off. The pack supports your MCOB suitability record for mortgages and your demands-and-needs statement for protection. It doesn’t replace your own judgement or your compliance process, but it means the reasoning is written down as the case runs, not reconstructed afterwards.
Why it matters now
When clients can get a product comparison from an AI assistant in seconds, what they value is someone who understands their situation, tells them honestly what to do and stands behind it. The more of the adviser’s day goes on forms, the less of that they can offer. Clear, consistent files also help you show Consumer Duty outcomes across the whole client bank, not just in the cases a file checker happens to pick.
Checklist
- Pick two tasks from your Move 1 list. Document gathering and suitability letter drafting are good first choices because they are frequent and easy to check.
- Design your evidence pack template. Agree the headings with your compliance support: circumstances, objectives, options considered, reasons for the recommendation, reasons for rejecting alternatives, risks explained.
- Set a checking rule. The adviser confirms every figure and every lender or insurer criterion against its source before sign-off, and the pack records who checked it.
- Run a pilot on ten cases. Time them against your Move 1 baseline and review each pack with a file checker.
- Switch on always-on reviews for one group. Start with clients whose deals expire in the next nine months, and have AI draft a personal note for each one for the adviser to approve.
Illustrative example
Illustrative example: an adviser handles 40 cases a month and spends about 3 hours on each file write-up. With AI drafting the evidence pack as the case runs, write-up falls to about 1 hour of checking and editing. That saves 2 hours per case, or 80 hours a month (40 cases at 2 hours each). The adviser uses part of that time to call every client with a deal expiring in the next six months, before the lender’s own switch offer lands.
Common mistakes
- Letting AI choose the product. It shortlists and explains; the adviser makes the recommendation and owns it.
- Generic evidence packs. If every pack reads the same, it isn’t evidence of anything. Each one must reflect the client’s own words and circumstances.
- Automating contact without judgement. An always-on review prompt is a reason for the adviser to get in touch, not a message to send on autopilot.
Coaching your clients
Many clients will arrive having already asked an AI assistant about rates or cover. Welcome it. Ask what it told them, then walk through where their own circumstances differ, such as variable income or a health condition. Clients are also facing change at work, and a job move or a switch to self-employment affects both borrowing and cover. Building those questions into every review is part of helping them adapt.