Crypto STR Reporting: Wolfsberg and Basel on Reports FIUs Can Use

On 11 November 2026 the Basel Institute on Governance published “Reporting suspicious crypto activity: Bridging supervisory expectations and operational realities”, a workshop report prepared with the Wolfsberg Group, which announced it the next day. It summarises a breakout session on crypto STR reporting held on 15 September 2026 at the 10th Global Conference on Criminal Finances and Cryptoassets, where supervisors, financial intelligence units (FIUs), financial institutions and cryptoasset service providers (CASPs) discussed the practical problems of crypto-related suspicious transaction and activity reporting (STR/SAR) and how to make those reports more useful to the authorities that receive them.

The friction runs in both directions. Participants described large volumes of reports with little context, which may put pressure on filers and FIUs alike, and filings made because an alert fired, a question stayed open or the entity wanted to avoid criticism. The report’s comparison table also lists the opposite failure: reports that describe transactions without explaining the suspicion, or that carry missing or inaccurate identifiers or data entered manually in inconsistent formats. Out of that came eight recurring considerations and one question the room could not settle, which is how far a bank or CASP should go in tracing a customer’s historical cryptoasset activity.

None of this moves the legal line. The report describes itself as an informal summary that represents no common position of participants, has not been adopted, endorsed or approved by their institutions, and should not be read as defining the legal threshold for reporting or as legal, regulatory, supervisory or compliance guidance. The filing test still comes from local law: national rules transposing Article 33 of Directive (EU) 2015/849 in the EU today, Article 69 of Regulation (EU) 2024/1624 (AMLR) from 10 July 2027, section 330 of the UK Proceeds of Crime Act 2002 for money laundering and section 21A of the Terrorism Act 2000 for terrorist-financing offences, the FinCEN suspicious activity report rules in the US and FINTRAC’s reasonable-grounds-to-suspect standard in Canada. The report is best read as a quality check laid over that test.

Related reading: Wolfsberg Non-Bank PSP Guidance: Banking the Sector Without De-Risking

The dates around the report

  • 15 September 2026: the two-hour breakout session runs at the 10th Global Conference on Criminal Finances and Cryptoassets, co-organised by the Basel Institute, Europol and UNODC and hosted by Luxembourg’s Bureau de gestion des avoirs (BGA).
  • 20 September 2026: AMLA’s public consultation on draft implementing technical standards (ITS) under Article 69(3) AMLR, the common EU format for reporting suspicions, closes. AMLA’s consultation page says results will follow.
  • 11 and 12 November 2026: the Basel Institute publishes the report and the Wolfsberg Group announces it.
  • 10 July 2027: the AMLR applies under its Article 90, including the Article 69 reporting duty, except for the obliged entities in Article 3, points (3)(n) and (o), which follow on 10 July 2029. AMLA’s guidelines on indicators of suspicious activity are due the same day under Article 69(5).
  • 10 July 2028: AMLA’s recommendations to FIUs on how to give feedback to reporting entities are due under Article 28(1) of Directive (EU) 2024/1640 (AMLD6).

For EU filers the operative date is 10 July 2027. Until then the reporting duty is whatever the national law transposing Article 33 of Directive (EU) 2015/849 says.

What the workshop report is, and what it can support

The session ran under the Chatham House Rule, so the report attributes views to “participants”, “several participants” or “some FIU representatives” and never to a named authority. That wording matters when the report is quoted internally: a procedure that says FIUs expect something, on the strength of this report, overstates what a moderated workshop summary can show.

The programme had three themes: what makes a crypto-related STR/SAR useful and actionable, the operational friction in preparing and assessing reports, and cooperation between public and private sectors.

The glossary at the end is more useful than most. Its entry for an alert reads as an automatic warning from a monitoring system that a transaction or customer looks unusual, with the rider that an alert alone does not mean the activity is suspicious; defensive reporting is filing mainly to avoid criticism or penalties; and a negative typology describes activity that may look suspicious but usually has a legitimate explanation. Those definitions can be lifted into an internal STR procedure with the report cited for the wording and the statutory test cited separately, since each of the eight considerations is stated to remain subject to the applicable legal and regulatory framework and to the reporting entity’s own risk-based assessment.

The eight considerations next to the binding rules

The executive summary lists eight considerations. The table shortens each one and adds, in the third column, binding provisions a reporting team would read alongside it. That third column is my mapping; the report itself cites no article of any law.

Consideration What participants said Binding text to read with it
1. Judge reports by intelligence value A good report is one authorities can act on; large volumes of low-context reports can strain both sides; a report of little value alone can matter when combined with other information Directive (EU) 2015/849 Art 33(1) and AMLR Art 69(1): report “regardless of the amount involved”
2. Alerts and red flags are a starting point Professional judgement decides; file when the information supports a genuine suspicion AMLR Art 69(2) on what a suspicion is based on
3. Explain the reasoning Why the activity is suspicious, what was reviewed, which steps were taken, how the evidence led to the conclusion FINTRAC “Details of suspicion” section; AMLD6 Art 28(1) feedback criteria
4. Get the data right and machine-readable Accurate identifiers, complete transaction data and structured formats such as XML, covering cryptoasset accounts AMLR Art 69(3) common format (draft ITS consulted in 2026); Regulation (EU) 2023/1113 Arts 18 and 22
5. Use analytics alongside other sources Analytics support human analysis; combine with CDD, monitoring, behavioural indicators and open sources AMLR Art 75(4) for information-sharing partnerships
6. Transparent analytics and common standards Standards on governance, taxonomy and explainability; limited support for full standardisation No binding standard among the instruments cited here
7. Join up the picture through cooperation Banks see fiat flows, CASPs see platform and on-chain activity, authorities see across reports AMLR Arts 69(8) and 75; tipping-off under Directive (EU) 2015/849 Art 39 and AMLR Art 73
8. Feedback improves reporting Case-by-case feedback is hard; typologies, negative typologies, thematic guidance and periodic quality feedback help AMLD6 Art 28(1): FIU feedback at least once a year

Rows 6 to 8 depend on parties outside the reporting entity, and the executive summary closes by calling the remaining work a joint task for public authorities and reporting entities.

One of the most consistent themes of the workshop was the distinction between unusual activity and suspicious activity. Red flags, alerts and risk scores are described as a starting point for analysis. Reporting entities are expected to apply professional judgement, assess the information available and decide whether it gives rise to a genuine suspicion. In the report’s comparison table, the less useful report is one filed solely because an alert was generated, or solely because information remains incomplete, without an assessment of whether the applicable reporting threshold is met. The more useful one is filed where, after an appropriate assessment, the applicable legal threshold for suspicion is met.

Participants were candid about how hard the call can be, describing a tension between timeliness and completeness. Entities can be criticised for reporting too late, while reporting too early may result in submissions holding little more than an initial alert. The tension was sharpest in smaller retail cases, where the question is how much investigative effort is proportionate before a decision. No common position emerged on the reporting threshold or on the point at which uncertainty should turn into an STR/SAR.

Local law fills that gap, and the wording differs more than the shared vocabulary suggests:

Regime Legal test Amount threshold Timing
EU today: national law transposing Directive (EU) 2015/849, Art 33(1) Knows, suspects or has reasonable grounds to suspect that funds are the proceeds of criminal activity or are related to terrorist financing None; attempted transactions included Promptly
EU from 10 July 2027: AMLR Art 69(1) Same test, applied to funds or activities and to links with criminal activity; suspicions arising from inability to conduct CDD must be reported None Promptly; replies to FIU requests within 5 working days unless the FIU shortens or extends that
UK regulated sector: Proceeds of Crime Act 2002, s.330 (money laundering), and Terrorism Act 2000, s.21A (terrorist financing) Under POCA s.330, knows, suspects or has reasonable grounds for knowing or suspecting that another person is engaged in money laundering, subject to the statutory conditions and defences. Under Terrorism Act 2000 s.21A, the corresponding test concerns another person committing or attempting an offence under sections 15 to 18. Where both money laundering and terrorist-financing suspicions arise, UKFIU guidance requires two separate SARs. None stated in s.330 As soon as practicable after the relevant information comes to the person. Under POCA section 330, disclosure is to a nominated officer or a person authorised by the Director General of the NCA; under Terrorism Act section 21A, disclosure is to a constable or nominated officer.
US banks: 31 CFR 1020.320 Knows, suspects or has reason to suspect one of the listed conditions At least $5,000 30 calendar days from initial detection; up to 30 more where no suspect is identified
US money services businesses covered by 31 CFR 1022.320(a)(1): the MSB categories in 31 CFR 1010.100(ff)(1) and (3) to (7), excluding businesses operating solely as check cashers Same wording, plus use of the business to facilitate criminal activity At least $2,000 ($5,000 for issuers reviewing money order or traveler’s check clearance records) 30 calendar days from initial detection
Canada: FINTRAC STR guidance Reasonable grounds to suspect that a transaction, including an attempted transaction, relates to the commission or attempted commission of a money laundering, terrorist activity financing or sanctions-evasion offence, under section 7 of the Proceeds of Crime (Money Laundering) and Terrorist Financing Act No monetary threshold As soon as practicable after completing the measures that establish reasonable grounds

The report’s line that filing “because questions remain open” adds little value needs care at this point. From 10 July 2027, Article 69(1) AMLR states that suspicions arising from the inability to conduct customer due diligence shall be reported. My reading is that the workshop’s point targets questions left open inside a file where due diligence was completed and no suspicion formed. Where required customer due diligence cannot be completed, Article 21(1) AMLR requires the obliged entity to refrain from the relevant transaction or relationship, terminate an existing relationship where applicable, and consider reporting to the FIU. Article 69(1) requires a report where that inability gives rise to suspicion; inability to complete due diligence does not itself remove the need to assess whether suspicion exists. The decision record should distinguish the failure to complete CDD from the grounds for any resulting suspicion.

The UK wording cuts the same way. Section 330 includes an objective limb, “reasonable grounds for knowing or suspecting”, so in my reading a no-file decision has to survive the question of what the person ought to have inferred from the information in front of them. FINTRAC’s guidance gives the clearest published description of the band in between. Simple suspicion, a hunch with no supporting facts, sits below the threshold and could prompt further assessment of related transactions. Reasonable grounds to believe, which rests on verified facts, is beyond what an STR requires.

For EU teams preparing the move from national transpositions to Article 69, our guide to the EU AML package and the countdown to 10 July 2027 covers the wider AMLR, AMLD6 and AMLA timetable.

What an actionable crypto STR narrative contains

The workshop’s description of a useful report is specific. It explains why the reporting entity considers the activity suspicious, which investigative steps were taken, what information was reviewed and how the evidence informed the final assessment. It tells a coherent story an analyst can follow, keeps to information relevant to the concern with enough context for further analysis, and is updated as new information emerges. The less useful version describes transactions or reproduces monitoring-tool output without explaining the suspicion, presents disconnected facts, or includes everything available regardless of relevance.

Regulators that publish STR guidance describe the same failure. FINTRAC lists, among common deficiencies, a narrative that does not elaborate on the grounds for suspicion or link them to the transactions in the report, and it asks for the facts, context and indicators that led to reasonable grounds to suspect. The guidance also sets a test of articulation: the suspicion should be explained so that another person with similar knowledge, experience or training would likely reach the same conclusion from the same information.

For EU filers, Article 69(2) AMLR supplies a skeleton for the narrative. A suspicion is to be based on the characteristics of the customer and their counterparts, the size and nature of the transaction or activity, its methods and patterns, the link between several transactions, the origin, destination or use of funds, and any other circumstance known to the obliged entity, including consistency with the due diligence information and the customer’s risk profile. Writing a crypto narrative under those headings forces the analyst to say what the on-chain flow means against the declared profile, the step that goes missing when monitoring output is pasted in unexplained.

Two things a narrative can leave out. Recital 138 AMLR says the predicate offence need not be specified where the obliged entity does not know it, and should be included where it is available. And the workshop was clear that relevant context matters more than including every piece of available information; an export of every hop an analytics tool can draw does not substitute for reasoning.

The report gives one worked example. A transaction was first flagged because of an indirect link to an address associated with terrorist financing concerns; open-source research then identified further indicators linked to extremist activity, and the combination of analytics, internal information and external research significantly strengthened the intelligence value of the report. The indirect exposure started the review, and the open-source research added the indicators that made the report stronger.

Reports also age. Private-sector participants described groups where different functions sit behind internal information barriers, so relevant facts emerge gradually and earlier reports need updating as understanding develops. The mechanism depends on the FIU. FINTRAC, for example, allows a submitted STR to be modified to add missing information or make corrections, with an explanation for the change, and Article 69(1) AMLR requires prompt answers to FIU requests for additional information.

Wallet addresses, identifiers and the structured-format question

Data quality emerged, in the report’s words, as one of the most consistent themes of the workshop. Participants said the usefulness of a report depends as much on the quality, completeness and structure of its information as on the underlying suspicion. Machine-readable formats work best when they capture consistent information on all relevant persons, entities, transactions and products, including bank accounts, payment accounts, virtual IBANs and cryptoasset accounts.

Several participants highlighted the benefits of XML reporting generated directly from internal systems in place of manual entry into portals: fewer manual errors, more complete reports, automated processing at the FIU and a better chance of linking reports that look unrelated. The cross-border argument carries particular weight for crypto, since the same customer or counterparty may be reported by different institutions in different jurisdictions at different times, and in some jurisdictions XML reporting supports automated dissemination through cross-border FIU cooperation mechanisms.

The report’s glossary notes that many FIUs use XML for structured STR submissions, for example through the goAML system developed by UNODC; our walkthrough of the goAML reporting workflow in Luxembourg covers how one FIU runs those submissions. Outside goAML, FINTRAC’s STR form already carries virtual currency address, reference and account fields in both the starting and the completing action of each transaction.

The EU’s answer to the format problem is still a draft. AMLA consulted from 2 July to 20 September 2026 on draft ITS under Article 69(3) AMLR specifying the format for reporting suspicions and for providing transaction records. The consultation paper describes a common core data set with templates adapted to different types of obliged entities, and separate transaction-record templates covering banking and payment activities, money remittance, crypto-asset accounts and correspondent services.

For suspicion-report formats, AMLA proposes a two-phase process. FIUs would assess the draft data points during the first two years after publication of the ITS, followed by a coordinated review of the annexes and technical implementation under deadlines that remain unspecified. Transaction-record formats have a separate implementation timetable, also left undetermined in the consultation draft. The proposed two-year FIU assessment phase should therefore not be applied automatically to transaction-record reporting.

File format stays with the FIUs under that draft. The draft ITS specifies data points and their attributes without mandating a single EU technical language, file format or XML schema. However, proposed Article 3(8) requires suspicion reports to be transmitted electronically in a machine-readable format through FIU reporting platforms, and proposed Article 4(3) requires machine-readable transaction records in formats specified by FIUs. Both provisions allow specified exceptions where normal reporting channels or formats cannot be used. These requirements remain proposals until the ITS is adopted and applicable. A project plan that budgets for an EU-mandated XML schema is planning for something the draft does not contain.

Crypto filers in the EU already hold one structured data source the narrative can draw on. Under Articles 18 and 22 of Regulation (EU) 2023/1113, the beneficiary’s CASP and any intermediary CASP must take missing or incomplete originator or beneficiary information into account as a factor when assessing whether a transfer of crypto-assets, or any related transaction, is suspicious and whether it is to be reported to the FIU. A factor in the assessment is the same status the workshop gives an alert. Our note on FATF travel rule implementation and the enforcement gap covers the state of travel rule rollout across jurisdictions.

Blockchain analytics: one input with known failure modes

Participants valued blockchain analytics for visibility into flows, links between entities and network analysis, and for many institutions it has become an integral part of crypto investigations. The cautions were concrete. Activity that first looked suspicious was sometimes explained once more information arrived, and in some cases unusual patterns traced back to incomplete wallet attribution inside the analytics tool, with nothing suspicious in the customer’s behaviour. Providers use different methodologies, clustering techniques and data sources, which can occasionally produce different conclusions about the same activity.

The report’s comparison table marks a report that treats an analytics result as the sole basis for suspicion as less useful. Behavioural analysis carried weight in the examples discussed: money-mule networks and coordinated activity were identified from transaction timing, clustering of bank accounts and customers, IP addresses, VPN usage and links to over-the-counter trading. Detection based only on on-chain activity was described as insufficient.

On attachments, the report’s position is conditional. Because authorities may not have the same analytics platforms, datasets or attribution methods as the filer, some FIU representatives suggested attaching relevant analytics reports or extracts, and the executive summary frames that as something that may help where the reporting framework permits it and where it is useful and proportionate. Whether a national reporting system accepts attachments, and in what form, depends on that FIU’s own instructions.

Many participants wanted more transparency about how results are generated, including methodologies, attribution approaches and confidence levels, and generally favoured common standards on governance, taxonomy and explainability, with limited support for full standardisation of models or datasets. The practical consequence, as I see it, is that the explainability burden stays with the filer for now: a narrative that states what an attribution rests on, the confidence the provider attaches to it and whether other sources corroborate it is easier for an FIU without the same tool to weigh.

Binding EU law applies the same principle in a narrower place. Under Article 75(4) AMLR, members of a partnership for information sharing may share information generated by artificial intelligence, machine learning or algorithms only where those processes were subject to adequate human oversight, may not rely solely on information received through the partnership, and must assess such information before taking decisions that affect the business relationship. Those conditions apply to partnership sharing from 10 July 2027. Read alongside the workshop, they point the same way: algorithmic output enters a decision after a person has reviewed it.

How far back to trace historical crypto activity

The question the workshop could not settle was how far reporting entities should be expected to evidence and trace historical cryptoasset activity, and how much residual uncertainty should count in a suspicion assessment. It affects banks and CASPs alike. Investigations can require review of activity spanning multiple years and numerous exchanges, intermediaries, self-hosted wallets and other blockchain transactions, sometimes through decentralised services or counterparties that no longer operate.

The executive summary calls the reference case typical: a bank receives funds from a MiCA-licensed CASP, the broad source of wealth is understood, the customer’s explanations are consistent and there are no obvious red flags. The report’s illustration follows the chain outwards. The licensed CASP keeps records; behind it may sit another exchange in a different jurisdiction, a self-hosted wallet whose owner is hard to identify, a DeFi service with no central company and a closed exchange whose records may be lost. The possibility and accuracy of tracing fade with each hop.

The involvement of a regulated provider was generally viewed as an important element of the risk assessment. Several participants cautioned against assuming that all licensed providers have the same AML/CFT maturity, supervisory effectiveness or implementation standards, so a licence under Regulation (EU) 2023/1114 (MiCA) was treated as relevant to the extent of further enquiries without being decisive. The glossary applies the same logic to any regulated provider: being subject to regulation is relevant to the assessment of AML/CFT risk and to the extent of due diligence, but does not determine either.

The most usable output of the report’s tracing discussion is a split between two sources of uncertainty, with two drivers that can fall on either side:

Driven by risk indicators Driven by practical limits Can be either
Inconsistent customer explanations Old or missing records Activity across multiple jurisdictions
Links to high-risk addresses Platforms no longer operating Customers reluctant to share wallets
Unexplained transaction patterns

Proportionality in information gathering was the related open point. Some cryptoasset investors and entrepreneurs are reluctant to disclose lists of wallet addresses, historical wallets or detailed holdings for privacy and personal security reasons. Participants asked how far financial institutions can reasonably go in requesting such information and whether unwillingness to provide it should itself be treated as a risk indicator. The report records the question without answering it, and notes that many participants would value more clarity on supervisory expectations, proportionality and the level of evidence reasonably expected.

Until that clarity arrives, a documented position is the defensible one. My working approach would record, for each historical-tracing decision, where tracing stopped, why it stopped, and whether the residual uncertainty sits in the risk-indicator column or the practical-limits column. The information requests themselves carry a tipping-off risk: FINTRAC’s guidance says reporting entities should not request information they would not normally request during a transaction if they believe doing so would tip off the client.

Cooperation and feedback between banks, CASPs and FIUs

The visibility split is simple to state. Banks see fiat flows, customer profiles and broader account behaviour; CASPs see platform and on-chain activity; authorities see across reports. Several participants described the interface between traditional finance and the crypto ecosystem as one of the areas where information asymmetries are most pronounced, and saw closer engagement between banks and CASPs as a way to close gaps within legal, operational and confidentiality limits.

In the EU, Article 75 AMLR sets the terms for one form of that engagement from 10 July 2027. Article 75 permits strictly necessary information sharing within qualifying partnerships, subject to the information categories in paragraph 3 and the safeguards in paragraph 4. Under Article 75(4)(f), sharing is confined to customers associated with higher risks identified in the specified EU or national risk assessments, customers falling within the situations identified in Articles 29 to 31 or 36 to 46 AMLR, or customers for whom further information is needed to determine whether they present higher risk. The remaining conditions include prior supervisory verification of the partnership arrangements, a data protection impact assessment before processing personal data, recording each exchange, each entity’s own assessment of transactions and FIU agreement before sharing information on reported suspicious transactions. Where a partnership’s activities produce a reportable suspicion, Article 69(8) AMLR permits the relevant obliged entities to designate one participant to submit a report identifying the participating entities and their contact details, or to report separately with a reference to the partnership.

Where the reporting entities are established in several Member States, the information must be reported to every relevant FIU, with submission arranged through an obliged entity established in the Member State of each receiving FIU. A single submission to only one Member State’s FIU does not satisfy that cross-border requirement. Outside the permitted routes, the tipping-off prohibition applies: Article 39 of Directive (EU) 2015/849 today and Article 73 AMLR from 10 July 2027.

US money services businesses work under a different rule. Where several of them are involved in one transaction, the obligation to report rests with each, but no more than one report is required if the report filed contains all relevant facts (31 CFR 1022.320(a)(4)). Our coverage of FATF public-private partnerships for sharing AML data gives the international picture on information-sharing arrangements.

Feedback was the other recurring theme. Many reporting entities want to know how their reports are used and which information is most valuable, while authorities pointed to the difficulty of case-specific feedback with rising volumes and limited resources. The middle ground the report describes includes negative typology publications, thematic guidance, periodic quality feedback and structured dialogue, plus one model in which reporting entities assign an initial risk assessment to each report and the FIU later gives periodic feedback on timeliness, completeness and overall quality. The report names the Global Conference, the Wolfsberg Group and the Europol Financial Intelligence Public-Private Partnership (EFIPPP) as forums for that work.

EU law turns part of this into an obligation on FIUs once Member States transpose AMLD6, which is due by 10 July 2027. Article 28(1) AMLD6 requires Member States to ensure FIUs give obliged entities feedback on their reporting of suspicions at least once a year, individually or to groups or categories of entities, covering at least the quality of the information, the timeliness of reporting, the description of the suspicion and the documentation provided. The same article says the feedback is not to be understood as covering each report, and that it is also made available to supervisors for risk-based supervision. Because supervisors receive it too, a pattern of thin narratives can surface in supervision as well as at the FIU.

Testing a crypto STR workflow against the report

The report is short enough to use as a review script. Five checks follow from it:

  1. Escalation records separate the alert, the concern and the suspicion, name the legal test applied (the national transposition of Article 33 of Directive (EU) 2015/849 or Article 69 AMLR, section 330 POCA for UK money laundering or section 21A of the Terrorism Act 2000 for UK terrorist financing, 31 CFR 1020.320 or 1022.320 for the covered US institutions, or section 7 of Canada’s PCMLTFA and FINTRAC’s reporting standard), and, where a file closed with open questions, say whether the question blocked customer due diligence.
  2. The narrative template follows the Article 69(2) AMLR basis for suspicion, or FINTRAC’s facts, context and indicators, and explains what the on-chain flow means against the customer’s profile.
  3. Wallet addresses and crypto-asset account identifiers go into structured fields wherever the FIU’s schema provides them, with travel rule gaps recorded as an assessment factor under Articles 18 and 22 of Regulation (EU) 2023/1113.
  4. Any analytics result cited states what the attribution rests on and the confidence the provider gives, records whether other sources corroborate it, and is attached only where the FIU’s framework permits.
  5. Historical-tracing decisions record the stopping point and whether the residual uncertainty comes from risk indicators or practical limits.

Checks 2 to 4 can be tested on a sample of filed reports. Checks 1 and 5 also need the files that ended without a report, since that is where the workshop’s open questions sit.

Frequently Asked Questions

Can an MLRO cite the workshop report to justify a decision not to file?

It can support the reasoning, for example the distinction between an alert and a suspicion, but the record has to name the statutory test that was applied, because the report’s own disclaimer rules out reading it as the legal threshold. A no-file record that cites the report and no statutory test is weak on its face.

A customer will not disclose historical wallet addresses for privacy reasons. Is that reportable?

The workshop left open whether that reluctance should itself be treated as a risk indicator. In my reading the answer turns on what the information was needed for. If refusal prevents required due diligence from being completed, Article 21(1) AMLR requires the entity to take the prescribed steps in relation to the transaction or business relationship and to consider reporting to the FIU. Article 69(1) requires reporting where the inability to conduct CDD gives rise to suspicion. Both provisions generally apply from 10 July 2027. If due diligence is complete and the request went beyond it, the refusal is one input to the overall assessment.

An incoming crypto transfer arrives without the required originator information. Does that trigger an STR?

Not automatically. Under Regulation (EU) 2023/1113, the beneficiary’s CASP applies risk-based procedures to decide whether to execute, reject, return or suspend the transfer (Article 17) and treats the missing information as a factor when assessing whether the transfer or a related transaction is suspicious and reportable to the FIU (Article 18). Where a counterparty CASP repeatedly fails to provide the information, Article 17(2) requires that failure and the steps taken to be reported to the AML/CFT competent authority, which is a separate report from an STR to the FIU.

Our bank and our CASP sit in the same group behind information barriers. Can the suspicion be shared between them?

From 10 July 2027, Article 73(3) AMLR allows disclosure between obliged entities in the same group, and with their branches and subsidiaries in third countries, where those branches and subsidiaries fully comply with group-wide policies and procedures, including procedures for sharing information within the group under Article 16, and those policies comply with the AMLR. Until then the national transposition of Article 39 of Directive (EU) 2015/849 sets the exceptions to the tipping-off prohibition. The workshop’s point about information emerging gradually inside groups is also a reason to update an earlier report when the other entity’s facts arrive.

A crypto withdrawal is pending while we decide. Can it be executed?

Under Article 35 of Directive (EU) 2015/849, obliged entities refrain from carrying out transactions they know or suspect to be related to the proceeds of criminal activity or to terrorist financing until they have reported and have complied with any further specific instructions from the FIU or the competent authorities under national law, and inform the FIU immediately afterwards where refraining is impossible or likely to frustrate efforts to pursue the beneficiaries. From 10 July 2027, Article 71 AMLR keeps that rule and adds that the entity may carry out the transaction, after assessing the risks of proceeding, if it has received no instruction to the contrary from the FIU within 3 working days of submitting the report.

We are a CASP established in one Member State serving customers in others. Which FIU receives the report?

The compliance officer transmits it to the FIU of the Member State where the obliged entity transmitting the information is established: Article 33(2) of Directive (EU) 2015/849 today and Article 69(6) AMLR from 10 July 2027. Article 53(1) of Directive (EU) 2015/849 already requires that FIU to forward promptly a report concerning another Member State to the FIU there; once AMLD6 is transposed (the transposition deadline is 10 July 2027), Article 31(1) AMLD6 carries the same duty and allows the report, or the relevant information obtained from it, to be forwarded. Article 31(3) AMLD6 gave AMLA a mandate to draft regulatory technical standards on when a report concerns another Member State.

Does a MiCA market abuse report replace the AML report on the same activity?

They are separate reports with different recipients. Article 92 of Regulation (EU) 2023/1114 requires persons professionally arranging or executing cryptoasset transactions to report reasonable suspicions of market abuse without delay to the relevant national competent authority. Commission Delegated Regulation (EU) 2025/885, in force since 9 September 2025, establishes the operative STOR arrangements and reporting template. ESMA’s December 2024 final report records the consultation-paper position that STORs should not cover scams, payment fraud or account takeover without an identifiable connection with market abuse. An AML suspicion goes to the FIU. One set of facts can trigger both; our guide to MiCAR STOR reporting under Article 92 covers the market abuse side.

Key Takeaways

  • EU filers switch from national Article 33 rules to Article 69 AMLR on 10 July 2027; escalation templates that cite the legal test need updating for that date.
  • Treat an analytics risk score as an investigative input and assess the underlying transactions, attribution evidence and other available information against the applicable legal threshold for suspicion. Seek corroboration where appropriate, but do not defer a required report solely because an additional information source has not confirmed an already substantiated suspicion.
  • Record the stopping point and the type of residual uncertainty in every historical-tracing file; there is no common standard yet for where tracing should stop.
  • Scope AMLA common-format work around data-point mapping, FIU-specific technical formats and the proposed machine-readable submission requirements. The draft’s two-year FIU assessment phase concerns suspicion-report formats; transaction-record formats have a separate, currently unspecified implementation timetable.
  • Plan EU bank-CASP information sharing from 10 July 2027 through Article 75 AMLR partnerships, which need supervisor verification and a data protection impact assessment before the first exchange and may share information only on the customers described in Article 75(4)(f).
  • Under Article 28(1) AMLD6 as transposed, FIUs must provide feedback on reporting quality at least annually, but that feedback may be addressed to individual obliged entities or to groups or categories of entities. The feedback must also be made available to supervisors.

Sources and References

  • Basel Institute on Governance and The Wolfsberg Group, “Reporting suspicious crypto activity: Bridging supervisory expectations and operational realities”, workshop report, November 2026 (PDF): baselgovernance.org
  • Basel Institute on Governance, publication page, 11 November 2026: baselgovernance.org/resources/publications/crypto-str-report
  • The Wolfsberg Group, news item, 12 November 2026: wolfsberg-group.org
  • Directive (EU) 2015/849 (Articles 33, 35, 39 and 53): EUR-Lex CELEX 32015L0849
  • Regulation (EU) 2024/1624 (AMLR), including Articles 69, 71, 73, 75 and 90 and Recital 138: EUR-Lex CELEX 32024R1624
  • Directive (EU) 2024/1640 (AMLD6), Articles 28 and 31: EUR-Lex CELEX 32024L1640
  • Regulation (EU) 2023/1113 on information accompanying transfers of funds and certain crypto-assets, Articles 17, 18 and 22: EUR-Lex CELEX 32023R1113
  • Regulation (EU) 2023/1114 on markets in crypto-assets (MiCA): EUR-Lex CELEX 32023R1114
  • AMLA, consultation on draft ITS on the format for reporting suspicions and providing transaction records (Article 69(3) AMLR): amla.europa.eu
  • AMLA, consultation paper on the draft ITS under Article 69(3) AMLR, 2 July 2026 (PDF): amla.europa.eu
  • AMLA, factsheet “A common EU format for reporting suspicions”, 2 July 2026 (PDF): amla.europa.eu
  • AMLA, press release launching the consultation, 2 July 2026: amla.europa.eu
  • Cellule de renseignement financier (Luxembourg), notice on the AMLA consultation and its 20 September 2026 deadline: crf.public.lu
  • Commission Delegated Regulation (EU) 2025/885 (market abuse arrangements and STOR template under MiCA Article 92(2)): EUR-Lex
  • ESMA, final report on draft technical standards under MiCA Article 92(2) (ESMA75-453128700-1278), December 2024: esma.europa.eu
  • ESMA, consultation paper on MiCA market abuse RTS (ESMA75-453128700-1002), March 2024: esma.europa.eu
  • Proceeds of Crime Act 2002, section 330: legislation.gov.uk
  • Terrorism Act 2000, section 21A: legislation.gov.uk
  • UKFIU, SARs Best Practice Guidance, Chapter 2: Submitting a SAR: nationalcrimeagency.gov.uk
  • 31 CFR Part 1020 (section 1020.320, reports by banks of suspicious transactions), 2025 annual edition: govinfo.gov
  • 31 CFR Part 1022 (section 1022.320, reports by money services businesses of suspicious transactions), 2025 annual edition: govinfo.gov
  • Proceeds of Crime (Money Laundering) and Terrorist Financing Act, section 7: laws-lois.justice.gc.ca
  • FINTRAC, “Reporting suspicious transactions to FINTRAC”: fintrac-canafe.canada.ca

Before Article 69 AMLR applies on 10 July 2027

What the report offers crypto-exposed banks and CASPs is a shared description, from supervisors, FIUs and filers in the same room, of a report an FIU can use; the historical-tracing question remains open. For EU filers the next fixed date is 10 July 2027, when Article 69 AMLR becomes the directly applicable reporting rule and AMLA’s guidelines on indicators of suspicious activity fall due, with the common suspicion-report format following its proposed two-phase timetable once the ITS is adopted. The record to have in place before then is an escalation note for each crypto case that names the legal test applied and documents where historical tracing stopped and why.

Disclaimer: The information on RegReportingDesk.com is for educational and informational purposes only. It does not constitute legal, regulatory, tax, or compliance advice. Always consult your compliance officer, legal counsel, or the relevant supervisory authority for guidance specific to your institution.

Similar Posts