AML Compliance Statistics 2026
By Axis Intelligence Research
Co-author: Sarah Davis | Last updated: August 11, 2026 | License: CC BY 4.0
U.S. financial institutions filed 4.8 million Suspicious Activity Reports in fiscal year 2025, the highest volume on record. In the same period, 2,795 FBI investigative subjects tied to Bank Secrecy Act reports were convicted. According to Axis Intelligence Research, that is 58.2 documented convictions per 100,000 reports filed.
Quick Answer
U.S. SAR filings reached 4.8 million in FY2025 and Currency Transaction Reports reached 21.5 million, from roughly 335,000 registered filers, per FinCEN. Global AML, KYC, sanctions and CDD penalties totalled $3.8 billion in 2025, a second consecutive annual decline. The FATF grey list stood at 22 jurisdictions as of 19 June 2026.
Key Findings
- According to Axis Intelligence Research, U.S. SAR volume grew at a 3.73% compound annual rate between FY2022 and FY2025, while Currency Transaction Report volume grew at 1.44% over the same window.
- The UK Financial Intelligence Unit denied a record £382.6 million to suspected criminals through Defence Against Money Laundering requests in 2024-25, against 866,616 SARs received, per the National Crime Agency.
- According to Axis Intelligence Research, five refused DAML cases out of 2,059 produced 45.8% of all UK refused-DAML asset denial value in 2024-25.
- FinCEN valued the two-year delay of the investment adviser AML rule at approximately $1.45 billion in avoided pro forma compliance cost, per the final rule published at 91 FR 36.
- According to Axis Intelligence Research, the AML Signal Yield Ratio reads 236.3 for the United Kingdom in 2024-25 and 58.2 for the United States in FY2025, computed on each jurisdiction’s own outcome definition.
How Many Suspicious Activity Reports Are Filed Each Year?
Reporting volume is the load-bearing number in every AML compliance statistics discussion, because everything else — headcount, alert queues, technology spend, regulator expectations — scales off it.
FinCEN’s Year in Review for Fiscal Year 2025 puts SAR filings at 4.8 million for the year ending 30 September 2025, up from 4.7 million in FY2024, 4.6 million in FY2023 and 4.3 million in FY2022. Currency Transaction Reports came in at 21.5 million, recovering from a dip to 20.5 million the year before.
| BSA report type | FY2022 | FY2023 | FY2024 | FY2025 | Source |
|---|---|---|---|---|---|
| Suspicious Activity Reports | 4.3M | 4.6M | 4.7M | 4.8M | FinCEN |
| Currency Transaction Reports | 20.6M | 20.9M | 20.5M | 21.5M | FinCEN |
| CMIRs | 128,100 | 143,200 | 152,100 | 132,400 | FinCEN |
| Form 8300 cash reports | 431,800 | 421,500 | 470,400 | 420,400 | FinCEN |
| FBARs | 1.5M | 1.6M | 1.7M | 1.8M | FinCEN |
Source: FinCEN Year in Review for Fiscal Year 2025, retrieved 11 August 2026.
The two headline series are diverging. According to Axis Intelligence Research, SAR filings compounded at 3.73% a year across FY2022-FY2025 while CTRs compounded at 1.44% — SAR growth running roughly 2.6 times faster than the cash-threshold series it sits alongside. Over the full three years that is +11.6% for SARs against +4.4% for CTRs.
That gap matters because the two reports are generated by different machinery. A CTR is a mechanical trigger: cash over $10,000, file it. A SAR is a judgment call produced by a monitoring model, an alert queue, and an analyst. When the judgment-based series pulls away from the mechanical one, the driver is not more cash moving — it is more detection logic pointed at more products.
Reporting Density Per Institution
FinCEN reports approximately 335,000 registered financial institutions and other filers. According to Axis Intelligence Research, that works out to 14.3 SARs and 64.2 CTRs per registered filer per year — a figure that is almost meaningless as a benchmark for any individual bank, and useful precisely because it shows how skewed the distribution is. Large depository institutions file in the tens of thousands; most registered filers file nothing in a given year. The same concentration shows up on the payments side, as our contactless payments statistics work documents.
For a fuller breakdown by industry type across calendar years 2014 to 2024, FinCEN publishes downloadable filing trend data through its SAR Filings by Industry tables.
Sarah Davis, Axis Intelligence Research: The filing-volume record gets reported every year as evidence the system is working harder. It is evidence the system is filing more. Those are different claims, and the second one is the only one the data supports on its own. Everything that follows in this article is an attempt to find out what the 4.8 million turn into.
What Happens to a Suspicious Activity Report After It Is Filed?
This is the question almost no AML statistics page answers, because the outcome data sits in different documents from the filing data and uses different units.
BSA reporting is queried heavily. FinCEN records over 2.52 million searches through BSA Search by 464 federal, state and local agencies in FY2025, plus more than 61.38 million searches by federal partners with integrated access — 63.9 million searches in total. According to Axis Intelligence Research, that is 2.43 database queries for every SAR and CTR filed that year.
Queries are not outcomes. The outcome figures FinCEN publishes are narrower:
| FY2025 outcome measure | Value | Source |
|---|---|---|
| FBI subjects arrested tied to BSA reports | 5,957 | FinCEN |
| FBI subjects indicted tied to BSA reports | 4,727 | FinCEN |
| FBI subjects convicted tied to BSA reports | 2,795 | FinCEN |
| IRS-CI cases searched against BSA data | 94% | FinCEN |
| IRS-CI investigations originating from a BSA filing | 11.7% | FinCEN |
| Conviction rate, IRS-CI cases with associated BSA filings | 98% | FinCEN |
Source: FinCEN Year in Review for Fiscal Year 2025, retrieved 11 August 2026.
The 11.7% figure deserves more attention than it gets. BSA data touches nearly every IRS-CI case — 94% are searched against it — but originates only about one in nine investigations. The regime works mostly as corroboration and enrichment for cases opened another way, not as a lead generator. Both of those are legitimate functions. They imply very different answers to “is the filing volume proportionate,” which is why the distinction should be in every effectiveness debate and rarely is.
Where BSA Reporting Bites Hardest
| FBI case program | Active investigations linked to SARs or CTRs | Source |
|---|---|---|
| Organized Crime Drug Enforcement | 47.6% | FinCEN |
| Complex Financial Crime | 38.8% | FinCEN |
| Transnational Organized Crime and FTO | 32.8% | FinCEN |
| Public Corruption | 27.4% | FinCEN |
| International Terrorism | 15.1% | FinCEN |
Source: FinCEN Year in Review for Fiscal Year 2025, retrieved 11 August 2026.
Drug enforcement, not terrorism financing, is where the reporting regime does most of its work. Nearly half of active OCDETF investigations link back to a SAR or CTR, against 15.1% for international terrorism — a spread of more than three to one between the crime type the regime is most associated with in public discourse and the one it actually feeds.
Introducing the AML Signal Yield Ratio (ASYR)
According to Axis Intelligence Research, the missing number in AML compliance statistics is not how much is filed or how much is fined. It is how much reported suspicion converts into a documented outcome. So we built a metric for it.
ASYR — AML Signal Yield Ratio. Documented enforcement or asset-denial outcome events recorded per 100,000 suspicion reports filed in the same reporting period, computed inside a single jurisdiction using that jurisdiction’s own published outcome definition.
Formula
ASYR = (documented outcome events / suspicion reports filed, same period) x 100,000
Baseline readings, as of 11 August 2026
| Jurisdiction | Period | Outcome events (numerator) | Reports filed (denominator) | ASYR |
|---|---|---|---|---|
| United Kingdom | Apr 2024 – Mar 2025 | 2,048 AFOs, forfeitures and restraints obtained through DAMLs | 866,616 SARs received | 236.3 |
| United States | Oct 2024 – Sep 2025 | 2,795 FBI subjects convicted tied to BSA reports | 4,800,000 SARs filed | 58.2 |
Inputs: National Crime Agency UKFIU SARs Annual Report 2024-25; FinCEN Year in Review FY2025. Both retrieved 11 August 2026.
Why the two readings are published side by side and never merged. The UK numerator counts asset-denial events; the U.S. numerator counts convicted persons. Those are not the same unit, and averaging them would produce a number that means nothing. Axis Intelligence Research treats ASYR as a within-jurisdiction time series — the question it answers is “is this regime converting reported suspicion into outcomes at a better or worse rate than last year,” not “which country is better at AML.” Any league table built by comparing the two readings would be misuse of the metric, and we will say so publicly if one appears.
Precision note. The U.S. denominator is published by FinCEN rounded to 4.8 million, so the U.S. reading carries roughly one significant figure of denominator precision. The UK reading is exact on both sides. Readings are baseline values, not a series with history: neither can yet be described as a high or a low.
Sarah Davis, Axis Intelligence Research: The interesting thing about ASYR is not the level, it is the denominator’s behaviour. A compliance function under pressure files more. If outcome events do not move with filings, the ratio falls, and a falling ratio is the clearest quantitative signal that defensive filing has taken hold. That is the number an MLRO should be watching, and no supervisor currently publishes it.
How Much Money Does the AML System Actually Deny Criminals?
The UK publishes the cleanest asset-denial data of any major jurisdiction, and 2024-25 was a record year.
The National Crime Agency’s SARs Annual Report 2024-25 records £382.6 million denied to suspected criminals through Defence Against Money Laundering requests, up 59% from £240.1 million the year before. The three-year rolling average — which the UKFIU publishes precisely because single years swing on individual cases — sits at £298.4 million.
| UK SARs and DAML metrics | 2023-24 | 2024-25 | Source |
|---|---|---|---|
| SARs received | 872,048 | 866,616 | NCA UKFIU |
| DAML requests received | 57,081 | 57,666 | NCA UKFIU |
| DAML requests refused | 2,881 | 3,276 | NCA UKFIU |
| Funds denied from DAML requests | £240.1M | £382.6M | NCA UKFIU |
| AFOs, forfeitures and restraints via DAMLs | 1,785 | 2,048 | NCA UKFIU |
| Average DAML decision turnaround | 2.8 days | 3.11 days | NCA UKFIU |
Source: National Crime Agency, UKFIU SARs Annual Report 2024-25, retrieved 11 August 2026.
The Concentration Nobody Reports
The record headline hides an extreme distribution. Refused DAMLs produced £359,884,586 of asset denial across 2,059 cases in 2024-25. The UKFIU’s own banding table shows one case in the £50M-plus band worth £103,000,000, and four cases in the £10M-£50M band worth £62,003,845.
According to Axis Intelligence Research, five cases — 0.243% of all refused-DAML asset denial cases — produced £165,003,845, or 45.8% of the total refused-DAML denial value.
| Refused DAML value band | Cases 2024-25 | Total value 2024-25 | Source |
|---|---|---|---|
| £0 – £500K | 1,983 | £100,267,020 | NCA UKFIU |
| £500K – £1M | 37 | £25,329,846 | NCA UKFIU |
| £1M – £5M | 31 | £49,598,375 | NCA UKFIU |
| £5M – £10M | 3 | £19,685,500 | NCA UKFIU |
| £10M – £50M | 4 | £62,003,845 | NCA UKFIU |
| £50M+ | 1 | £103,000,000 | NCA UKFIU |
| Total | 2,059 | £359,884,586 | NCA UKFIU |
Source: National Crime Agency, UKFIU SARs Annual Report 2024-25, retrieved 11 August 2026.
This is why year-on-year percentage changes in “funds denied” should be treated with suspicion by anyone quoting them. Remove the single £103 million case and the 2024-25 total falls to £256.9 million, which is a good year rather than a record one. The UKFIU says as much in its own footnotes; almost nobody who cites the 59% increase says it too.
Sector Distribution of UK Reporting
| Sector | Share of SAR Portal submissions 2024-25 | Source |
|---|---|---|
| Banking | 85.54% | NCA UKFIU |
| Financial services | 10.34% | NCA UKFIU |
| Virtual assets | 1.67% | NCA UKFIU |
| Gambling | 0.88% | NCA UKFIU |
| Accountancy | 0.71% | NCA UKFIU |
| Legal | 0.40% | NCA UKFIU |
| Property | 0.10% | NCA UKFIU |
| High value goods | 0.05% | NCA UKFIU |
Source: National Crime Agency, UKFIU SARs Annual Report 2024-25 (848,086 SARs submitted via the SAR Portal), retrieved 11 August 2026.
Banking carries 85.54% of UK reporting. Property — the sector named in essentially every national risk assessment as a laundering channel — accounts for 0.10%. The UKFIU explicitly declines to comment on whether sector volumes are proportionate to risk, leaving that to supervisors. Axis Intelligence Research will note the arithmetic: estate agents and letting agents produced roughly one SAR for every 855 filed by banks.
What Is the Recovery Rate on Reported Fraud Proceeds?
FinCEN’s Rapid Response Program gives the closest thing to a clean recovery cascade in AML statistics, because the same document publishes the money in, the money frozen, and the money returned.
| Rapid Response Program, FY2025 | Value | Source |
|---|---|---|
| Reported to the programme | $362.6M | FinCEN |
| Frozen by partners | $182.2M | FinCEN |
| Returned to victims | $95.5M | FinCEN |
Source: FinCEN Year in Review for Fiscal Year 2025, retrieved 11 August 2026.
According to Axis Intelligence Research, that is a 50.2% freeze rate on reported funds, a 52.4% return rate on frozen funds, and a 26.3% end-to-end recovery rate — roughly one dollar returned for every four reported. Business email compromise dominated the frozen total at $90.95 million, ahead of investment fraud at $39.86 million and account takeover at $21.12 million. Consumer exposure to these same typologies is tracked in our financial literacy statistics dataset.
A 26.3% end-to-end rate is far better than most fraud-recovery benchmarks and far worse than the impression created by headlines about funds frozen. Both halves of that sentence are the point. The cascade only works when the report arrives fast enough for a foreign FIU to act, which is why the referral count — 392 in FY2025 — matters more than the dollar total for anyone designing a payments fraud response. Our analysis of online scam statistics covers the victim-side view of the same pipeline.
How Big Are AML Fines in 2025 and 2026?
Global penalties fell for a second consecutive year. Fenergo, which compiles an annual dataset of enforcement actions across EMEA, North America and Asia Pacific, put 2025 AML, KYC, sanctions and CDD penalties at $3.8 billion, down from $4.6 billion in 2024 and $6.6 billion in 2023.
| Enforcement measure | 2023 | 2024 | 2025 | Source |
|---|---|---|---|---|
| Global AML/KYC/sanctions/CDD penalties | $6.6B | $4.6B | $3.8B | Fenergo |
| North America change year on year | — | — | -58% | Fenergo |
| EMEA change year on year | — | — | +767% | Fenergo |
| APAC change year on year | — | — | +44% | Fenergo |
Source: Fenergo annual enforcement findings, published 13 January 2026; retrieved 11 August 2026.
According to Axis Intelligence Research, that is a 42.4% decline from the 2023 peak. The regional split is the story: U.S. regulators issued $1.676 billion (44.1% of the global total) and French authorities $1.11 billion (29.2%), the latter driven by a single $985 million penalty against a Swiss bank — the largest AML-related penalty of the year.
A documented source disagreement. Fenergo’s own report landing page describes 2025 as a year in which regulators issued “over $4 billion” in AML penalties, while its January 2026 release states $3.8 billion for AML, KYC, sanctions and CDD combined. The press release also attributes a 58% decline to North America and, in a quoted passage, a 61% decline to the United States specifically. Separately, figures circulating from Norton Rose Fulbright commentary put 2024 global AML fines at $10.4 billion — nearly 2.3 times the Fenergo 2024 figure of $4.6 billion.
Axis Intelligence Research will not reconcile these into a single number. The variance is definitional: what counts as an AML fine, whether deferred prosecution agreements and forfeiture amounts are included, and whether the figure is per-action or per-institution. Any AML compliance statistics page presenting one global fines total without stating its scope is presenting a number whose meaning it cannot defend. We publish the range and the reason instead.
Sarah Davis, Axis Intelligence Research: Falling fines are being read in board packs as falling enforcement risk. Read the composition instead. North America down 58% on a shutdown-and-staffing story, EMEA up 767% on the conclusion of investigations opened years ago. Enforcement value is a lagging indicator of supervisory attention by roughly the length of an investigation, which in the UK case just proved to be the COVID period. The pipeline argument is the only one that survives contact with the regional data.
Which Countries Are on the FATF Grey List in 2026?
The FATF grey list is the single most operationally consequential list in country-risk scoring, and it changed at the June 2026 plenary.
Per the FATF statement on Jurisdictions under Increased Monitoring dated 19 June 2026, Bosnia and Herzegovina and Iraq were added; Algeria and Namibia were removed. The list stands at 22 jurisdictions. The high-risk list subject to a call for action remains Iran, North Korea and Myanmar.
The Grey List Is Not One Category — Axis Segmentation
Treating 22 jurisdictions as a single risk tier throws away most of the information in the statement. According to Axis Intelligence Research, the 19 June 2026 text supports five distinct states, based on what the FATF says about each country’s progress:
| State | Count | Jurisdictions | Basis in the 19 June 2026 statement |
|---|---|---|---|
| Cleared for on-site assessment | 4 | Bulgaria, Côte d’Ivoire, DR Congo, Monaco | FATF made an initial determination that the action plan is substantially complete |
| Active implementation | 10 | Angola, Bolivia, Bosnia and Herzegovina, Iraq, Kenya, Lao PDR, Lebanon, Nepal, Venezuela, Virgin Islands (UK) | Progress noted, deadlines running |
| Deadlines expired | 4 | Cameroon, Haiti, South Sudan, Vietnam | FATF states all action plan deadlines have expired with work remaining |
| Reporting deferred | 2 | Kuwait, Papua New Guinea | Jurisdictions chose to defer; February 2026 statements carried forward |
| Frozen pending on-site access | 2 | Syria, Yemen | Action plans completed at technical level; on-site visit not possible for security reasons |
Source: Axis Intelligence Research segmentation of the FATF statement, Jurisdictions under Increased Monitoring, 19 June 2026; retrieved 11 August 2026.
Two observations follow that no other published source makes. First, four of the 22 are already through the substantive work and awaiting a site visit — a near-term exit pipeline that a country-risk model treating grey-list status as static will misprice. Second, the Syria and Yemen statements in force on 19 June 2026 are dated February 2023, meaning the two jurisdictions have carried the same three-year-old text through nine plenaries. Whatever those listings signal, it is not current supervisory assessment.
The FATF states that grey-list status does not call for enhanced due diligence as a class and does not envisage cutting off entire customer categories. That guidance is routinely ignored in practice, which is why the segmentation above is worth building into risk scoring rather than applying a flat listing flag.
What Changes for AML Compliance in the EU in 2026 and 2027?
The EU is mid-transition to its first centralised AML supervisor, and the dates are now fixed.
The Authority for Anti-Money Laundering and Countering the Financing of Terrorism (AMLA), based in Frankfurt, will directly supervise 40 of the most complex high-risk financial institutions or groups in the EU from 2028, with the first selection round running in 2027. In December 2025 AMLA published draft regulatory technical standards on risk assessment and on selection, establishing — for the first time — a single methodology that supervisors across all 27 member states will use to assess money laundering and terrorist financing risk.
| EU AML transition milestone | Date | Source |
|---|---|---|
| AMLA operational, Frankfurt | 1 July 2025 | AMLA |
| Draft RTS on risk assessment and selection published | 18 December 2025 | AMLA |
| Selection round for directly supervised entities | 2027 | AMLA |
| Direct supervision of 40 entities begins | 2028 | AMLA |
Source: AMLA, press release of 18 December 2025; retrieved 11 August 2026.
The number that matters for most firms is not 40. It is 27 — the count of national supervisors who will be assessing risk with the same data points and the same criteria. Firms that have quietly relied on the softest supervisor in their footprint lose that arbitrage before the first directly supervised entity is even named.
What Does AML Compliance Cost? The Only Government Figure Worth Quoting
AML compliance cost estimates are dominated by vendor surveys with undisclosed methodologies. There is one exception, and it is buried in a U.S. rulemaking.
When FinCEN delayed the investment adviser AML rule from 1 January 2026 to 1 January 2028, it was required to quantify the effect. The final rule published at 91 FR 36 on 2 January 2026 estimates the two-year delay reduces pro forma costs by approximately $1.45 billion, with a net present value of $1,453.63 million at a 7% discount rate and $1,523.60 million at 3%, and annualised savings of $183.01 million at 7%.
The underlying figures are more useful than the headline. FinCEN’s original analysis projected aggregate sector compliance expenditure of $800 million in 2026 and $780 million in 2027, in 2022 dollars. According to Axis Intelligence Research, that is $1.58 billion of first-two-year AML program cost for a single sector, removed from the schedule by one rulemaking.
| Investment adviser AML rule, cost parameters | Value | Source |
|---|---|---|
| Original effective date | 1 January 2026 | FinCEN / Federal Register |
| Revised effective date | 1 January 2028 | FinCEN / Federal Register |
| Projected sector expenditure, 2026 | $800M (2022 dollars) | FinCEN / Federal Register |
| Projected sector expenditure, 2027 | $780M (2022 dollars) | FinCEN / Federal Register |
| Estimated pro forma cost avoided | ~$1.45B | FinCEN / Federal Register |
| Registered investment advisers, year-end 2024 | 15,870 | FinCEN / Federal Register |
| Exempt reporting advisers, year-end 2024 | 5,743 | FinCEN / Federal Register |
| Comment letters received | 22 | FinCEN / Federal Register |
Source: Federal Register, 91 FR 36, published 2 January 2026; retrieved 11 August 2026.
FinCEN acknowledges in the same document that delay may prolong exposure to illicit finance risks it previously identified in its 2024 Investment Adviser Risk Assessment, and that several transparency organisations opposed the delay on national security grounds. The rulemaking is unusual in publishing both the cost saved and the reason the saving is contested. For the technology side of the same cost equation, see our analysis of AI in banking statistics.
Sarah Davis, Axis Intelligence Research: $800 million in year one across roughly 21,600 advisers is the closest thing the industry has to an audited per-firm AML startup cost, and it came out of a regulatory impact analysis rather than a survey. Vendors quoting compliance cost figures should be asked which of the two their number resembles.
Why the Most-Quoted Money Laundering Statistic Should Be Retired
Nearly every AML article opens with the same figure: 2% to 5% of global GDP is laundered annually, or $800 billion to $2 trillion.
The UNODC money laundering overview states that range, describing the dollar figures as being in “current US dollars.” Its lineage runs back to a February 1998 speech by then-IMF Managing Director Michel Camdessus to the FATF plenary, in which 2% to 5% of global GDP was offered as a probable consensus range, with the explicit caveat that accuracy could not be guaranteed.
According to Axis Intelligence Research, the dollar range has a testable internal property. If $800 billion is 2% of global GDP, the implied GDP base is $40 trillion. If $2 trillion is 5%, the implied base is also $40 trillion. Both bounds resolve to the same figure, which means the dollar range is anchored to a global economy of roughly $40 trillion — a base last consistent with the early 2000s, not with 2026.
The percentage range and the dollar range are therefore not two ways of saying the same thing. Either the percentages hold and the dollar figures are badly stale, or the dollar figures hold and the percentages have fallen. The two cannot both be current. Axis Intelligence Research recommends citing the percentage range with its 1998 provenance stated, and treating any dollar figure derived from it as an artefact.
What the Filing-Sharing Data Says About Collaboration
Section 314(b) of the USA PATRIOT Act lets financial institutions share information with each other to identify money laundering. Uptake remains thin.
| Section 314(b) participation, FY2025 | Value | Source |
|---|---|---|
| Registered financial institutions | 7,200+ | FinCEN |
| Institutions referencing 314(b) in SAR narratives | 1,690+ | FinCEN |
| SARs referencing 314(b) | 65,480+ | FinCEN |
| Terrorism-related SARs referencing 314(b) | 80+ | FinCEN |
Source: FinCEN Year in Review for Fiscal Year 2025, retrieved 11 August 2026.
According to Axis Intelligence Research, 314(b)-referencing SARs represent 1.36% of the FY2025 SAR total, and only 23.5% of registered participants actually referenced the provision in a narrative. Registration is not participation. Against 335,000 registered BSA filers overall, 314(b) registration itself covers around 2% of the filing population. Information-sharing reluctance has a well-documented parallel in security disclosure, which we cover in our cloud security statistics analysis.
Section 314(a), the law-enforcement-initiated channel, ran at 1,065 requests in FY2025 — 1,007 money-laundering-related and 58 terrorism-related — covering 6,883 subjects and generating 64,259 positive responses across 13,887 financial institutions.
Methodology
Collection. Every figure in this article was retrieved from a primary source document opened during production on 11 August 2026. Sources: FinCEN Year in Review for Fiscal Year 2025 (PDF); FinCEN SAR Filings by Industry; National Crime Agency UKFIU SARs Annual Report 2024-25 (PDF); FATF Jurisdictions under Increased Monitoring, 19 June 2026; AMLA press release of 18 December 2025; Federal Register 91 FR 36; UNODC money laundering overview; Fenergo annual enforcement findings of 13 January 2026. No figure in this article is drawn from memory, from a secondary aggregator, or from an estimate produced to fill a gap.
Reporting periods. U.S. figures are federal fiscal years ending 30 September. UK figures are financial years ending 31 March. Global penalty figures are calendar years. These are not aligned and have not been forced into alignment.
ASYR formula. ASYR = (documented outcome events / suspicion reports filed in the same period) x 100,000, computed within a single jurisdiction using that jurisdiction’s own published outcome definition. UK numerator: 2,048 AFOs, forfeitures and restraints obtained through DAMLs. UK denominator: 866,616 SARs received. U.S. numerator: 2,795 FBI investigative subjects tied to BSA reports convicted. U.S. denominator: 4,800,000 SARs filed. Readings published 11 August 2026 are baseline values.
Other derived figures. Compound annual growth rates use (end / start)^(1/n) – 1 over three years. Concentration shares divide named band values by the published total. Recovery rates divide each stage of the Rapid Response cascade by the stage above it. The implied GDP base divides each UNODC dollar bound by its corresponding percentage bound. All arithmetic was recomputed in Python before publication and every derived row is flagged axis_calculated = yes in the dataset with its formula in method_note.
Scope and caveats. ASYR readings are not comparable across jurisdictions because outcome definitions differ; the U.S. denominator is published rounded to 4.8 million, which caps that reading’s precision. Global penalty totals reflect one compiler’s inclusion rules and are documented above as disputed rather than reconciled. UK asset-denial totals are dominated by individual large cases, which is why the three-year rolling average is reproduced alongside the annual figure. FATF segmentation reflects the state of each jurisdiction as described in the 19 June 2026 statement and will change at the next plenary. The UNODC range is a 1998-vintage estimate and is presented as such.
About This Dataset
File: aml-compliance-statistics-2026.csv Rows: 138 observations Coverage: United States, United Kingdom, European Union, global; fiscal and calendar years 2022 to 2026 Licence: CC BY 4.0 Columns: metric, value, unit, as_of_date, geography, segment, source_org, source_document, source_url, retrieved_date, is_primary, axis_calculated, method_note, data_type, license
Every number appearing in this article exists as a row in the CSV with its source organisation, source document, source URL and retrieval date. Derived figures carry axis_calculated = yes and a formula in method_note. Axis-originated figures — the two ASYR readings, the FATF state segmentation, and the implied GDP base — carry data_type = original_metric.
Citation
APA: Axis Intelligence Research. (2026). AML Compliance Statistics 2026: Filings, Fines, and What the System Actually Recovers. Axis Intelligence. https://axis-intelligence.com/aml-compliance-statistics/
MLA: Axis Intelligence Research. “AML Compliance Statistics 2026: Filings, Fines, and What the System Actually Recovers.” Axis Intelligence, 2026, axis-intelligence.com/aml-compliance-statistics/.
Chicago: Axis Intelligence Research. “AML Compliance Statistics 2026: Filings, Fines, and What the System Actually Recovers.” Axis Intelligence. 2026. https://axis-intelligence.com/aml-compliance-statistics/.
Frequently Asked Questions
How many SARs were filed in the United States in 2025?
Financial institutions filed 4.8 million Suspicious Activity Reports in fiscal year 2025, the year ending 30 September 2025, according to FinCEN’s Year in Review. That is up from 4.7 million in FY2024. Currency Transaction Reports over the same period reached 21.5 million.
What is a good SAR-to-outcome conversion rate for an AML programme?
No supervisor publishes a benchmark, which is why Axis Intelligence Research built the AML Signal Yield Ratio. The two published baseline readings are 236.3 outcome events per 100,000 reports for the UK in 2024-25 and 58.2 for the U.S. in FY2025, on different outcome definitions. Institutions should track their own ratio over time rather than against another jurisdiction.
Does a rising SAR volume mean an AML programme is working?
Not on its own. SAR volume grew 11.6% between FY2022 and FY2025 while Currency Transaction Reports grew 4.4%, per FinCEN. Volume growth is consistent with better detection and with defensive filing. The distinguishing test is whether outcome events move with filings, which is what ASYR measures.
How many countries are on the FATF grey list right now?
Twenty-two, as of the FATF statement dated 19 June 2026. Bosnia and Herzegovina and Iraq were added at the June 2026 plenary; Algeria and Namibia were removed. Iran, North Korea and Myanmar remain on the separate high-risk list subject to a call for action.
Should grey-list exposure trigger enhanced due diligence automatically?
The FATF states it does not call for enhanced due diligence to be applied to grey-listed jurisdictions as a class, and that the standards do not envisage cutting off whole customer categories. Axis Intelligence Research segmentation of the June 2026 statement finds four of the 22 already cleared for on-site assessment and four with expired action-plan deadlines — states that a flat listing flag cannot distinguish.
When does AMLA start supervising banks directly, and which ones?
Direct supervision of 40 of the most complex high-risk financial institutions or groups begins in 2028, with selection carried out in 2027 under the common risk methodology AMLA published in draft in December 2025. Firms outside the 40 still face the same assessment criteria applied by their national supervisor.
Do investment advisers have AML programme obligations in 2026?
Not under the investment adviser AML rule. FinCEN delayed its effective date from 1 January 2026 to 1 January 2028 in a final rule published at 91 FR 36 on 2 January 2026. Advisers affiliated with banks or broker-dealers may still fall under existing obligations through those entities.
Why did AML fines fall in 2025 if enforcement is intensifying?
Global AML, KYC, sanctions and CDD penalties fell to $3.8 billion in 2025 from $4.6 billion in 2024, per Fenergo, but the regional composition inverted: North American penalties fell 58% while EMEA rose 767%. Penalty value lags supervisory attention by the length of an investigation, so annual totals describe cases opened years earlier rather than current scrutiny.
What proportion of UK suspicious activity reports come from outside banking?
Banking accounted for 85.54% of SARs submitted through the UK SAR Portal in 2024-25 and financial services a further 10.34%, per the National Crime Agency. Property firms accounted for 0.10% and high value goods dealers 0.05%, despite both sectors featuring prominently in national risk assessments.