FR-2026-0017

AI-assisted complaints: what the Ombudsman’s 54% finding means

New Legal Ombudsman research separates AI use from complaint quality. What the samples show, and how firms can keep the facts and requested remedy clear.

Analysis
1
October 6, 2026
1.0
Law
Communication
Mara Ellison
Editorial pen name
2
1
54%
AI use among surveyed people who made or considered a regulated-service complaint.
Original sources linked below; scope and limitations in the method.
54%
AI use among surveyed people who made or considered a regulated-service complaint.
478
The relevant survey subgroup, within a total sample of 2,196.
Original retained
A working summary should remain linked to the client’s original account.
Respond to the substance and requested remedy. Keep original messages, summaries, evidence and corrections connected.
The Formal Record is published by Letro, which has a commercial interest in formal communication software. AI assisted the research and drafting. Mara Ellison is an editorial pen name used by The Formal Record; it does not identify a separate staff journalist.
1.0 | 6 Oct 2026 | First published following the 29 September–5 October source scan.
legal-ombudsman-ai-complaints-2026-research

The brief

The Legal Ombudsman’s 1 October research announcement reports AI use by 54% of people who had made or considered a complaint about a regulated service in the previous year. That is not a finding that AI wrote 54% of all legal complaints. The study combines a public survey with separate reviews of complaint files.

Our take is to preserve the person’s account, the supporting evidence and the outcome they seek as distinct parts of the record. An AI-generated summary can help organise a complaint, but it should remain traceable to the original. A polished letter should neither receive extra credibility nor be dismissed simply because it appears machine-written.

The full record

A complaint may arrive as a fluent, formal letter and still leave the reader uncertain about what happened. The practical question is whether the firm can recover the underlying issue and respond to it. The Legal Ombudsman’s new research and guidance, announced on 1 October 2026, give firms reasons to focus on that task.

Two different kinds of evidence

The research report, pages 5–6, describes an Ipsos survey of 2,196 people aged 16–75 in England and Wales, conducted on 13–17 March 2026. Of these, 478 had made or considered a formal complaint about a regulated service during the previous year. The 54% figure applies to that subgroup. It covers help considering a complaint as well as making one, across regulated services rather than legal services alone.

Separately, LeO reviewed 100 randomly selected legal complaint files. It identified indicators of AI use in 51. Those indicators are an assessment method, not confirmed authorship. In that subset, reviewers judged the apparent impact positive in 17 cases, neutral in 16 and negative in 18. The report also contains a separate 100-file conveyancing review; the two exercises should not be collapsed into a population-wide rate. These figures describe the reviewed files, not a controlled test of whether AI causes better outcomes. Source: report, pages 7–9.

What the guidance asks firms to address

LeO’s service-provider guidance describes both the potential to clarify a complaint and the risk of exaggerated or inaccurate correspondence. It cautions against responding in a way that escalates the exchange. Its consumer guide asks people to check facts, protect personal information and focus on a practical resolution.

Our analysis: make the complaint recoverable

A useful complaint record should allow a colleague to answer four questions: what service was expected, what the client says went wrong, which evidence supports or disputes that account, and what outcome the client wants. Keep the original message even when a summary is prepared. Link each material statement in the working summary to a document, message or confirmed account.

If the requested remedy is unclear, ask a short clarifying question. Record the answer beside the original request. Treat disputed facts as disputed, and identify who must investigate them. A long chain of replies can otherwise make an assumption look like an agreed fact.

For an internal trial, use fictional complaint files with deliberately conflicting dates or an unsupported legal reference. Ask reviewers to find the service issue, distinguish evidence from assertion and identify the unresolved questions. Measure omissions, corrections and the time needed for a second person to understand the file. Do not measure success only by how quickly a reply was drafted.

This is a proposed test, not a benchmark we have performed. The value of formal communication here is a record that keeps the client’s words, the firm’s findings and the eventual resolution connected. Our conveyancing workflow analysis applies the same test to instructions and approvals.

We read the research announcement, report and both guidance pages. Survey results and case-file assessments are separate evidence; apparent AI use is not proven authorship. We have not replicated the research or measured a product’s performance. The proposed internal trial is our analysis.

Brazil’s September fraud reports: why the denominator matters
brazil-certbr-september-2026-fraud-notifications
FR-2026-0029
Analysis
1
October 12, 2026
Law, Finance, Fiduciaries & trustees
Communication
CERT.br’s September figures show why a larger fraud share need not mean a similar rise in reports. Counts, reporting limits and a practical editorial check.
2
1
3,120
September notifications in CERT.br’s fraud category, not a count of victims.
CERT.br monthly table, September 2026; checked 12 October.
3,120
September notifications in CERT.br’s fraud category, not a count of victims.
Julian Vey
MAS’s public prospecting safeguards: record permission before the pitch
mas-public-prospecting-safeguards-october-2026
FR-2026-0028
Analysis
1
October 12, 2026
Finance
Communication
Singapore’s MAS plans stronger safeguards for public financial prospecting. Why the purpose of an approach and permission to continue need separate records.
2
1
Purpose first
Explain the intended commercial engagement before asking to proceed.
Original sources linked below; scope and limitations in the method.
Purpose first
Explain the intended commercial engagement before asking to proceed.
Julian Vey
iCapital’s adviser survey: confidence does not settle the client explanation
icapital-2026-adviser-survey-client-communication
FR-2026-0027
Analysis
1
October 12, 2026
Finance, Fiduciaries & trustees
Communication
iCapital’s 2026 adviser survey highlights client reporting and explanations of alternatives. What the evidence supports, and how to test investor understanding.
2
1
Reported priorities
The survey identifies perceived barriers rather than measured client outcomes.
Original sources linked below; scope and limitations in the method.
Reported priorities
The survey identifies perceived barriers rather than measured client outcomes.
Julian Vey
Six SRA alerts show how copied credentials support false instructions
sra-impersonation-alerts-6-12-october-2026
FR-2026-0026
Analysis
1
October 12, 2026
Law, Finance, Fiduciaries & trustees
Verification
Six SRA alerts dated 7–9 October cover false corporate, recovery and legal communications. A linked register explains the patterns and a separate late catch-up.
3
1
6 alerts
Six unique SRA warning URLs fall within the checked weekly period.
Original sources linked below; scope and limitations in the method.
6 alerts
Six unique SRA warning URLs fall within the checked weekly period.
Mara Ellison
AI agents and client instructions: UK and Singapore regulators sharpen the questions
ai-agents-client-instructions-ico-mas-october-2026
FR-2026-0025
Analysis
1
October 12, 2026
Law, Finance, Fiduciaries & trustees
Communication
The ICO seeks evidence on AI agents while MAS issues financial-sector AI guidelines. What these developments mean for permission, approval and client records.
2
1
Different status
The UK evidence call and Singapore final guidelines are separate developments.
Original sources linked below; scope and limitations in the method.
Different status
The UK evidence call and Singapore final guidelines are separate developments.
Mara Ellison
SRA’s competence review puts AI supervision inside everyday legal work
sra-competence-review-ai-supervision-2026
FR-2026-0024
Analysis
1
October 12, 2026
Law
Communication
The SRA proposes changes to solicitor competence on AI and supervision. A practical way to show who checked an AI-assisted document before it was sent.
2
1
Open consultation
The SRA is seeking views, rather than announcing a new rule.
Original sources linked below; scope and limitations in the method.
Open consultation
The SRA is seeking views, rather than announcing a new rule.
Mara Ellison
FCA’s 90-day fund proposal: a notice period is not a payment date
fca-90-day-fund-redemption-proposal-2026
FR-2026-0023
Analysis
1
October 12, 2026
Finance, Fiduciaries & trustees
Communication
The FCA proposes new redemption terms for funds holding illiquid assets. Why a 90-day notice period needs a clear explanation of dealing and payment dates.
2
1
Proposal
The consultation remains open until 11 December 2026.
Original sources linked below; scope and limitations in the method.
Proposal
The consultation remains open until 11 December 2026.
Julian Vey
FINMA’s digital identification update: what changes in November
finma-digital-identification-november-2026
FR-2026-0022
Analysis
1
October 12, 2026
Finance, Fiduciaries & trustees
Verification
FINMA’s revised digital identification circular starts in November. What changes, where a transition applies, and what an onboarding record should preserve.
2
1
1 November
The revised circular’s entry date requires a process-specific implementation check.
Original sources linked below; scope and limitations in the method.
1 November
The revised circular’s entry date requires a process-specific implementation check.
Julian Vey
FINMA’s Julius Baer findings: an endorsement is not evidence
finma-julius-baer-2026-client-risk-exceptions
FR-2026-0021
Analysis
1
October 6, 2026
Finance, Fiduciaries & trustees
Verification
FINMA’s Julius Baer findings show why a personal endorsement cannot replace source-of-wealth checks. How to record and revisit client-risk exceptions.
2
1
Reviewable
An exception needs a factual basis and a later review decision.
Original sources linked below; scope and limitations in the method.
Reviewable
An exception needs a factual basis and a later review decision.
Julian Vey
Singapore’s private banks target one-month account opening
singapore-private-banking-account-opening-one-month-target
FR-2026-0020
Analysis
1
October 6, 2026
Finance, Fiduciaries & trustees
Communication
Singapore’s private banking group aims to cut median account opening to one month. How to read the progress and measure the document exchanges behind it.
2
1
One month
PBIG’s target for member banks’ median account-opening times by end-2026.
Original sources linked below; scope and limitations in the method.
One month
PBIG’s target for member banks’ median account-opening times by end-2026.
Julian Vey
Fake job offers and payment support: check the claimed authority
impersonation-alerts-dubai-hong-kong-south-africa-october-2026
FR-2026-0019
Analysis
1
October 6, 2026
Law, Finance, Fiduciaries & trustees
Verification
Alerts from Dubai, Hong Kong and South Africa show different ways authority is claimed. What to verify before paying, sharing ID or trusting an offer.
3
1
USD 105
Certificate fee requested in the employment scam described by the DFSA.
Original sources linked below; scope and limitations in the method.
USD 105
Certificate fee requested in the employment scam described by the DFSA.
Julian Vey
AI solicitor recommendations need a separate status check
ai-solicitor-recommendations-regulatory-status-check
FR-2026-0018
Analysis
1
October 6, 2026
Law
Verification
A September study found gaps in AI solicitor recommendations. Its test design shows why a shortlist needs a dated check against the regulator’s own record.
2
1
30 test answers
The charity’s Doncaster repeat exercise; not a population error rate.
Original sources linked below; scope and limitations in the method.
30 test answers
The charity’s Doncaster repeat exercise; not a population error rate.
Mara Ellison
AI-assisted complaints: what the Ombudsman’s 54% finding means
legal-ombudsman-ai-complaints-2026-research
FR-2026-0017
Analysis
1
October 6, 2026
Law
Communication
New Legal Ombudsman research separates AI use from complaint quality. What the samples show, and how firms can keep the facts and requested remedy clear.
2
1
54%
AI use among surveyed people who made or considered a regulated-service complaint.
Original sources linked below; scope and limitations in the method.
54%
AI use among surveyed people who made or considered a regulated-service complaint.
Mara Ellison
FINMA consultation: record why information crosses a border
finma-direct-transmission-consultation-2026
FR-2026-0016
Analysis
1
October 5, 2026
Finance, Fiduciaries & trustees, Law
Communication
FINMA is consulting on direct information transfers to foreign authorities. The practical question is how to record purpose, authority and what was sent.
2
1
27 November
The consultation closes on 27 November 2026.
Linked original sources; reporting limits explained in the method.
27 November
The consultation closes on 27 November 2026.
Julian Vey
SJP and Partners Group: what a fund-change notice must explain
sjp-partners-group-fund-change-investor-notices
FR-2026-0015
Analysis
1
October 5, 2026
Finance, Fiduciaries & trustees
Communication
A proposed SJP fund closure and Partners Group restructuring raise different client questions. Clear notices must separate status, choices and liquidity.
3
1
Two events
A proposed closure and a proposed restructuring require different explanations.
Linked original sources; reporting limits explained in the method.
Two events
A proposed closure and a proposed restructuring require different explanations.
Julian Vey
Swiss ownership register opens: keep the evidence behind the entry
swiss-transparency-register-october-2026-client-evidence
FR-2026-0014
Analysis
1
October 5, 2026
Law, Finance, Fiduciaries & trustees
Verification
Switzerland’s ownership register opened on 1 October. What the launch means for document requests, client confirmations and the record behind an entry.
3
1
1 October
The Swiss transparency register began operating on 1 October 2026.
Linked original sources; reporting limits explained in the method.
1 October
The Swiss transparency register began operating on 1 October 2026.
Julian Vey
Nine SRA alerts: a genuine name cannot verify an instruction
sra-alerts-29-september-2-october-2026
FR-2026-0013
Report
1
October 5, 2026
Law, Finance, Fiduciaries & trustees
Verification
Nine SRA alerts cover deposit redirection, copied websites and recovery offers. Our linked register explains how to verify the specific instruction.
3
1
9
Distinct alerts dated 29 September–2 October in the SRA’s register.
Linked original sources; reporting limits explained in the method.
9
Distinct alerts dated 29 September–2 October in the SRA’s register.
Mara Ellison
Conveyancing in 2026: digital checks, fragmented communication
conveyancing-2026-digital-checks-client-communication
FR-2026-0012
Analysis
1
October 5, 2026
Law
Communication
InfoTrack’s 2026 index shows a gap between digital checks and client communication. What its figures measure, and how a firm can test its own workflow.
2
1
71%
Respondents reported phone, email or post for onboarding communication; vendor research.
Linked original sources; reporting limits explained in the method.
71%
Respondents reported phone, email or post for onboarding communication; vendor research.
Mara Ellison
September 2026: 48 SRA impersonation alerts
september-2026-sra-impersonation-alerts
FR-2026-0011
Report
1
September 30, 2026
Law, Finance, Fiduciaries & trustees
Verification
Our September register links 48 SRA scam alerts and explains what the count measures, the patterns it reveals and how firms can test client verification.
4
1
48
Distinct SRA alert URLs dated 1–30 September, reconciled on 5 October 2026.
SRA September index; 48 linked alerts; reconciled 5 Oct 2026.
48
Distinct SRA alert URLs dated 1–30 September, reconciled on 5 October 2026.
Mara Ellison
1
Revolut and Bitcoin Suisse: three questions about trust
revolut-bitcoin-suisse-september-2026-trust-controls
FR-2026-0010
Analysis
1
September 30, 2026
Finance
Verification
Revolut’s Swiss licence application, reported data disclosure and Bitcoin Suisse’s reorganisation raise separate questions about status, authority and access.
3
1
Application
Revolut’s announcement describes an application, not a granted Swiss banking licence.
Linked sources and reporting limits in the full record.
Application
Revolut’s announcement describes an application, not a granted Swiss banking licence.
Mara Ellison
Swiss wealth management: who owns the client instruction?
september-2026-swiss-wealth-client-instructions
FR-2026-0009
Analysis
1
September 30, 2026
Finance, Fiduciaries & trustees, Law
Communication
New Zurich offices, a boutique launch and DBS’s Swiss adviser model highlight a practical issue: who approves, receives and retains a client instruction?
3
1
Five
Distinct September developments, not a representative market survey.
Linked sources and reporting limits in the full record.
Five
Distinct September developments, not a representative market survey.
Julian Vey
Private markets: access, liquidity and the client record
september-2026-private-markets-investor-communication
FR-2026-0008
Analysis
1
September 30, 2026
Finance, Fiduciaries & trustees
Communication
New distribution routes and fund launches put investor communication under pressure. September’s reports show what a clear, retrievable explanation needs.
3
1
Access
New distribution does not remove eligibility, risk or liquidity conditions.
Linked sources and reporting limits in the full record.
Access
New distribution does not remove eligibility, risk or liquidity conditions.
Julian Vey
Legal technology in September: adoption needs evidence
september-2026-legal-technology-adoption-evidence
FR-2026-0007
Analysis
1
September 30, 2026
Law
AI
September’s legal technology reports show different ways to buy and use AI. The useful comparison is the quality, cost and record of completed work.
3
1
543
Respondents in the LexisNexis survey; reported adoption is not a measure of output quality.
Linked sources and reporting limits in the full record.
543
Respondents in the LexisNexis survey; reported adoption is not a measure of output quality.
Mara Ellison
Swiss investors: the AI–adviser trust gap
swiss-investors-ai-financial-adviser-trust
FR-2026-0006
Analysis
1
September 29, 2026
Finance
AI
A Swiss survey reports a narrow AI–adviser trust gap among Gen Z investors. The useful response is a clear way to check claims clients bring to their adviser.
3
1
February
The survey’s fieldwork month; the findings were published in September 2026.
IFZ/HSLU research article, 28 September 2026; linked in the record.
Snapshot
The reported age-group comparison does not establish a change in trust over time.
Sina Tadayon
FCA AML handover: can law firms retrieve the record?
fca-aml-supervision-law-firm-records
FR-2026-0005
Analysis
1
September 29, 2026
Law
Communication
The FCA plans to begin legal and accounting AML supervision in late 2028. A practical retrieval exercise can help firms examine their decision records.
4
1
Late 2028
When the FCA says it will begin taking on the additional AML supervision described in its speech.
FCA speech, 17 September 2026; linked in the record.
60,000
Entities across legal and accounting sectors in the FCA’s stated scope; not 60,000 law firms.
Mara Ellison
FCA disclosures: delivery is not understanding
fca-investor-disclosures-client-understanding
FR-2026-0004
Analysis
1
September 29, 2026
Finance
Communication
Only 6% of 132 investment disclosures met the FCA’s plain-English readability assessment. Here is a practical way to test what a reader has understood.
3
1
6%
Of 132 documents met the FCA’s plain-English readability assessment; not a measure of investor comprehension.
FCA disclosure review, 2 July 2026; linked in the record.
Documents
The 6% result describes a text assessment, not the share of clients who understood an investment.
Julian Vey
Aquila’s Wecan choice: testing Swiss data control
aquila-swiss-data-control-procurement
FR-2026-0003
Analysis
1
September 29, 2026
Finance, Fiduciaries & trustees, Public sector
Communication
Aquila’s Wecan selection and a Swiss government software study raise a useful buying question: what can an institution demonstrate about control of its data?
3
1
One file
Our proposed procurement test: follow a fictional client file through processing, review and export.
Formal Record analysis; source announcements linked below.
Selection
Wecan announced Aquila’s choice following a competitive process; implementation was underway.
Julian Vey
Brodies’ AI pilot: what law firms should measure
brodies-ai-pilot-law-firm-evaluation
FR-2026-0002
Analysis
1
September 29, 2026
Law
AI
Brodies chose Legora after a three-month, 180-person pilot. Our proposed scorecard helps law firms test quality, review time and readiness before buying.
3
1
180
Colleagues involved in Brodies’ pilot, according to the firm.
Brodies announcement, 25 September 2026; linked in the record.
3 months
Brodies reports testing Legora with legal and business support colleagues before choosing it.
Mara Ellison
SRA scam alerts: verifying client instructions
sra-scam-alerts-client-instruction-verification
FR-2026-0001
Analysis
1
September 29, 2026
Law
Verification
Sixteen SRA alerts in four days expose a practical question for law firms: can a client check who issued an instruction without trusting the message itself?
3
1
16
SRA alerts dated 22–25 September 2026; a publication count, not a count of victims.
SRA alert index; count and linked register in Method.
16
Alerts published over four days. This does not measure the frequency of fraud.
Mara Ellison
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.