AI consulting for FAP-licensed NZ financial advice firms, led by a CFP who runs a FAP-licensed practice in parallel. Compliance-first design, FMA-aware policy, clear advice-output safeguards, and pragmatic workflows for the adviser-adjacent tasks where AI is genuinely safe and useful.
Most AI consultants have never read the FAP code. We have. We run a FAP-licensed practice in parallel. The compliance framing isn't a side concern, it's the foundation.Craig Coupland · CFP · FSP 105424
The FAP licensing regime is now several years bedded in. The novelty has worn off; the operational reality of running a FAP-licensed practice in 2026 is steady, demanding, and increasingly squeezed. AI sits on top of that as either a relief valve or a new compliance headache, depending on how it's implemented.
Quarterly portfolio updates, KiwiSaver review cadence, regulatory disclosure refreshes, market-event commentary. The communication workload for a small advice firm with a few hundred clients is significant, and most of it isn't billable. AI can credibly cut drafting time on adviser-reviewed communications without crossing any advice-output line, when the workflow is designed properly.
Documentation requirements under the FAP code haven't lightened. Advice records, suitability documentation, conflict-of-interest disclosures, complaint handling, supervision logs. AI can accelerate parts of this (file note drafting from adviser-meeting recordings, structured first-pass advice records) without removing the adviser's professional responsibility for the substance.
Annual KiwiSaver reviews for a book of several hundred members involve significant repetitive document preparation, even when the underlying advice is straightforward. The structural pattern (read client info, summarise current position, compare against suitability framework, draft a review letter for adviser sign-off) is well-suited to AI assistance, again with the adviser's professional judgment preserved at the signoff step.
The FMA has publicly signalled it is paying attention to AI use across financial services, with specific concerns about offshore data residency and the risk that AI-generated material is treated by clients as advice when no advice was given. Firms with no written AI policy are in a weaker position at FMA monitoring than firms with documented governance, regardless of whether their actual practice is appropriate.
Most adviser practice-management platforms and CRMs are adding AI features in 2026. Many advisers are using these features (sometimes without fully realising) without an explicit firm-level approval. This is one of the most common compliance gaps we find: AI use that's already happened, hasn't been mapped against the FAP code, and is creating exposure the firm doesn't yet know about.
FAP-licensed firms get a different sequence from our other engagements. We start with the Risk + Compliance Review, not with audit or implementation. Once the firm has a defensible posture, then we can talk about workflows. Compliance-first isn't a slogan; it's a sequence.
The default first engagement for any FAP-licensed firm. AI use inventory, FAP code mapping, acceptable use policy, vendor data residency, adviser certification. FMA-ready posture.
After the compliance posture is set. Adviser-adjacent workflows only: research summarisation, file note drafting, KiwiSaver review document prep, client communication drafting. Adviser signoff at every output.
Vendor evaluation specifically against FAP-grade criteria. Data residency, advice-output suitability, contractual processing terms, integration with practice management.
Most appropriate for FAP firms. Annual policy review, new-tool approvals, FMA-readiness updates, supervision of the adviser certification register. Ongoing governance without hiring a full-time compliance specialist.
AI use inside a FAP-licensed practice intersects with several overlapping rule sets. The FAP code is the dominant frame, but the Privacy Act 2020, the FMA's broader expectations, and the firm's own disclosure documents all matter.
The FAP code doesn't prohibit AI use. It requires that advice be suitable, given with care, in the client's interests, and with appropriate disclosure. The practical implication is that the adviser remains professionally responsible for any output that touches the client, full stop. AI can prepare the ground (research, drafting, file notes) but cannot substitute for the adviser's judgment on suitability or fair dealing. Workflows that respect this distinction are entirely compatible with the code. Workflows that don't aren't.
Three categorical no-go uses we encounter regularly:
Several uses are clearly safe with the right policy and adviser-review framing:
Client financial information is personal information at its most sensitive. We help firms pick tools with NZ/AU data residency, contractual processing addenda appropriate for retail client financial data, and disclosure language for client engagement documents. The FMA's posture on disclosure is unforgiving: clients who don't know AI is involved have not consented to it, and that's a serious matter.
For an AI-assisted KiwiSaver review letter, an FAP-aligned audit trail looks like: source client data archived in the practice management system (as today), AI-generated draft archived alongside, named adviser's reviewed and edited final letter as the deliverable, with a brief sign-off note recording the adviser's review and the suitability framework applied. Functionally identical to how a firm already documents adviser-only work, with one extra archived artefact.
A 12-adviser NZ financial advice firm, FAP-licensed and FMA-regulated, asked us to map their existing AI use against the rules. We found four risk categories, recommended two vendor changes, drafted an FMA-ready acceptable use policy, and certified all 12 advisers on safe AI use.
The engagement took three weeks end to end and resulted in a board-ratified, externally reviewed AUP, two vendor switches with no productivity loss, and a 100% adviser certification rate. The firm is now under a Fractional AI Director retainer for annual policy refresh.
Read the full case study →Illustrative engagement · real client outcomes published with consent only
Yes, with the right framework. The FAP code of conduct doesn't prohibit AI use. It requires that any advice given meets the suitability, fair-dealing, and priority-of-client-interest tests, and that the adviser remains professionally responsible for it. AI is permissible for adviser-adjacent work (research summarisation, file note drafting, internal communications). It is not safe for advice-output without strict adviser-review safeguards.
Three categorical no-go uses. First, AI generating advice content sent directly to clients without adviser review and signoff. Second, AI used to assess suitability where the adviser hasn't independently confirmed the suitability conclusion. Third, public AI chat interfaces processing client-identifying information. Each of these crosses a FAP code line and creates clear regulatory and PI exposure.
Several uses are clearly safe with the right policy. Research summarisation of public information (product PDSs, market commentary). File note drafting from adviser-meeting recordings with adviser review. Internal communication drafting. Initial structuring of client letters with adviser review and edit before send. Each of these accelerates the adviser without substituting for the adviser's judgment.
The FMA has publicly signalled increasing attention to AI in financial services. Specific concerns include offshore data residency without disclosure and AI-generated material being treated by clients as advice when no advice was given. Firms with a documented acceptable use policy, certified advisers, and a defensible audit trail are in a much stronger position than firms operating on individual adviser judgment alone.
Board-ratified, externally reviewed where appropriate. Specific approved-tools list maintained by the compliance officer. Clear prohibited uses. Data residency requirements. Audit trail and record-keeping obligations. Mandatory disclosure language for engagement letters. Incident response protocol. Supervision and certification register. Annual review built in. Coupland Consulting delivers exactly this format under our Risk + Compliance Review service.
Yes. Craig is a Certified Financial Planner and runs a FAP-licensed advice firm in parallel (Wealth Health). The disciplines that govern Wealth Health's operations (advice records, suitability, supervision, disclosure, conflict-of-interest handling) are the same disciplines we apply to AI compliance design at Coupland Consulting. No translation tax, no learning curve. The two practices are deliberately separate brands, with no FAP-licensing overlap, but the same operating standards.
Three weeks is typical for a Risk + Compliance Review for a FAP-licensed firm. Week one is AI use inventory and risk categorisation. Week two is policy drafting and external counsel review. Week three is vendor changes and adviser certification. Many firms add a Fractional AI Director retainer to handle annual policy review and new-tool approvals on an ongoing basis.
Book a 30-min discovery call with Craig. CFP, FSP 105424, FAP-licensed practice in parallel. We'll talk through your situation in FAP-code language, not consultant-speak.