For Jersey's fund administrators and trust companies, 2026 is the year regulatory expectation stopped being abstract. The Island's 2024 MONEYVAL mutual evaluation set the direction, and the follow-up has now arrived in concrete form: a rewritten AML/CFT/CPF Handbook, thematic examination findings that name specific failure patterns, and a supervisor that increasingly asks not whether a control exists, but whether the firm can prove it operated, on request, from its own records.
None of this should surprise anyone who has read the JFSC's recent publications. But the pace has picked up, and the standard of proof has changed. This article sets out what changed, what the examiners found, and what a practical response looks like for a regulated administration business.
What changed in 2026
The starting point is the 2024 MONEYVAL mutual evaluation. The report's recommendations triggered a programme of follow-up supervisory activity: thematic examinations targeted at areas the assessors flagged, an expanded supervisory risk data collection, and a rewrite of the core rulebook. Individually, each item is manageable. Together, they change the day-to-day relationship between a supervised firm and its regulator. Three developments define the year so far.
The rewritten AML/CFT/CPF Handbook took effect on 31 May 2026. The JFSC published the updated consolidated Handbook following its 2025 consultations, with revised sections covering complex structures and enhanced criminal background checks, and with further amendments expected later in 2026 as the new risk-based MLCO framework beds in. The rewrite is more than a reorganisation. It consolidates what the JFSC expects a supervised person to be able to demonstrate, and it shifts responsibility onto firms to collect and maintain the records that prove compliance.
The AML scope framework for fund structures is being redrawn. The JFSC's revised guidelines on Article 36 of the Proceeds of Crime (Jersey) Law 1999 set out a three-gateway test (a Jersey nexus, a scheduled activity, and the "as a business" requirement) for determining who falls within the AML regime. The consultation proposed an effective date of 30 April 2026; that date was deferred and the guidelines are being taken forward within a wider package of legislative and Handbook reforms. The direction is settled even if the commencement date moved. Firms administering fund structures should be mapping every entity in each structure against the test now, not waiting for the final implementation notice.
The SAR thematic findings landed. On 27 May 2026 the JFSC published feedback from its 2025 thematic examination of suspicious activity reporting, an exercise prompted directly by recommendations in the MONEYVAL report. The findings are specific, and they read across well beyond the nine firms examined.
What the SAR thematic examination found
The JFSC examined nine supervised persons. Four had no findings. Five did, and those firms were required to submit formal remediation plans to their supervisor.
The recurring weaknesses were not exotic. They were basic hygiene failures in the machinery of reporting:
- Outdated policies and procedures. SAR-related documentation that had not kept pace with current legislation or with the firm's actual practice.
- Weak SAR registers and record-keeping. Internal reports logged inconsistently, registers missing key fields, and gaps between what the register said and what the underlying file showed.
- Unformalised MLRO and Deputy MLRO oversight. Reporting lines, deputisation arrangements and role-specific training that existed in practice but were not documented, evidenced or tested.
"The firms that came through cleanly were not the ones with the biggest compliance teams. They were the ones whose registers, records and reporting lines matched what their procedures said, and who could show it on the day."
Trust companies and TCSPs should not treat this as a funds-only exercise. The failure modes are generic to any regulated administration business: the SAR perimeter, the register, the P&P version and the officer's decision trail exist in every supervised firm, and the same examination questions apply.
What "evidence on demand" means operationally
The practical standard under the rewritten Handbook is evidence on demand. It is not sufficient that customer due diligence was performed, that a disclosure decision was taken, or that a policy was reviewed. The firm must be able to produce the record promptly: who did it, when, on what basis, and under which statutory or Handbook provision.
The JFSC's own supervisory risk data for the trust company business sector shows how much capacity this already consumes:
Read those figures together. Roughly one in ten employees in the sector works in compliance and risk, and more than half of all customer relationships carry enhanced due diligence obligations, with the refresh cycles, senior sign-offs and record-keeping that EDD implies. When the supervisor's standard moves from "have a control" to "evidence the control on request", the cost lands on exactly this population. Firms that still assemble evidence by reconstructing email chains at examination time are paying that cost twice: once in the daily workload and again in examination findings.
A practical readiness checklist
Against the 2026 baseline, four checks cover most of the ground the examiners walked:
None of these checks requires new technology or a larger team. They require the discipline of keeping the record at the moment the work is done, and a periodic test: pick a decision from six months ago and see how long it takes to produce the full trail. If the answer is measured in days, the firm has an examination finding waiting to be written.
Regulatory note: in the 2025 SAR thematic, the difference between a clean outcome and a remediation plan was rarely the substance of the judgement calls. It was whether the firm's records, registers and documented oversight arrangements could evidence those judgements when the examiners asked. Under the rewritten Handbook, that is the default expectation for every supervised person, not a thematic one-off.
Where controlled automation helps
The evidencing burden described above is, at root, a process problem: the work of preparing registers, tracking policy versions and assembling decision packs is repetitive, high-volume and unforgiving of gaps. That is the category of work where automation is appropriate, provided the division of labour is strict. The pattern that survives regulatory scrutiny is narrow: AI prepares, named officers decide, and the system records.
Concretely, that means software drafting register entries with the statutory citation attached for an officer to confirm, flagging policies whose mapped Handbook sections have changed since the last review, and assembling the evidence pack behind an EDD sign-off so the officer's decision is captured as a committed record rather than an email. This is the model CoreAdmin is built on: the preparation is automated, every decision belongs to a named individual, and every step is written to an audit trail at the time it happens. The outcome the firm cares about is simple to state: when the JFSC says "show us", the answer is a lookup, not a reconstruction project.
The 2026 supervisory cycle rewards firms that treat evidence as a by-product of doing the work, not a separate exercise performed before an examination. The Handbook rewrite, the thematic findings and the sector data all point the same way. The firms that respond by tightening their records now will find the rest of the decade considerably cheaper than the firms that respond after their first remediation plan.
Sources
- JFSC, Updated AML/CFT/CPF Handbook and upcoming changes
- JFSC, Implementation date for AML/CFT/CPF Handbook enhancements
- JFSC, Feedback statement: Schedule 2 Guidelines
- Comsure, JFSC 2025 SAR thematic examination feedback: what MLROs and compliance teams must do now
- Comsure, JFSC annual supervisory risk data reports for the trust company business sector