Predicate offences, tax crime and the proceeds-of-crime chain

When I ask a room of new tax-authority recruits what makes money "dirty," the answers are almost always instinctive rather than legal: it is money from drugs, from bribery, from something obviously wrong. The law is more precise than instinct, and the precision matters because it determines whether a prosecutor can charge laundering at all. The starting point in every modern AML statute is the requirement of a predicate offencePredicate offenceA specified underlying criminal offence whose proceeds are alleged to be the subject of a laundering charge. — a specified underlying crime whose proceeds are alleged to have been laundered. Money is not dirty in the abstract; it is dirty because it is traceable, directly or through substitution, to a particular category of criminal conduct that the legislature has designated as capable of generating launderable proceeds.

FATF does not leave this designation to individual states' unguided discretion. Its glossary sets out a list of designated categories of offencesDesignated categories of offencesFATF's glossary list of offence types every jurisdiction must cover, through an all-crimes or schedule-based approach, as launderable predicates. that every jurisdiction is expected to criminalise as predicates to money laundering, spanning participation in an organised criminal group, terrorism and terrorist financing, trafficking in human beings and migrant smuggling, sexual exploitation, illicit trafficking in narcotic drugs and psychotropic substances, illicit arms trafficking, illicit trafficking in stolen goods, corruption and bribery, fraud, counterfeiting currency, counterfeiting and piracy of products, environmental crime, murder and grievous bodily harm, kidnapping and hostage-taking, robbery or theft, smuggling, extortion, forgery, piracy, insider trading and market manipulation, and, since the 2012 revision I will return to in the next lesson; tax crimes. The list is deliberately broad and deliberately generic: FATF does not require that every jurisdiction use identical statutory language, only that each designated category be captured by at least one predicate offencePredicate offenceA specified underlying criminal offence whose proceeds are alleged to be the subject of a laundering charge. in domestic law, whether through a single all-crimes approachAll-crimes approachA predicate-offence model treating any criminal conduct, domestic or foreign, as capable of generating launderable proceeds. or a specific list.

That last phrase, all-crimes versus list-based, is probably the most consequential drafting choice a legislature makes in this field. An all-crimes approachAll-crimes approachA predicate-offence model treating any criminal conduct, domestic or foreign, as capable of generating launderable proceeds., which the United Kingdom, Australia and South Africa (through the broad "unlawful activity" definition in the Prevention of Organised Crime Act 121 of 1998, POCA) have all adopted in one form or another, treats any conduct that is a criminal offence, wherever committed, as capable of generating proceeds for laundering purposes. A list-based approach, more common in civil-law systems and in some early-generation AML statutes, ties the predicate to an enumerated schedule of offences, which is administratively cleaner but creates gaps whenever new criminal conduct, a novel cyber-fraud typology, for instance — emerges faster than the legislature amends the schedule. I tell officials working in list-based systems that their first diagnostic question on any new typology is whether it fits an existing scheduled item or falls into a gap that needs a law-reform submission, because a beautifully investigated case can still collapse if the predicate itself is not covered.

The Palermo Convention — the United Nations Convention against Transnational Organized Crime, adopted in 2000 and in force since 2003, is the treaty backbone beneath the FATF standard. Article 6 obliges states parties to criminalise the laundering of proceeds of crime and, critically, Article 6(2)(b) requires each state to apply the offence to "the widest range of predicate offences," extending at minimum to serious crime as defined elsewhere in the Convention (an offence punishable by a maximum deprivation of liberty of at least four years). Palermo's companion instrument for corruption-specific predicates is the United Nations Convention against Corruption (UNCAC, 2003), which performs the same function for bribery, embezzlement, trading in influence and illicit enrichment offences, and which additionally, through its Chapter V, establishes the framework for cross-border asset recovery that we will examine later in this course. Reading POCA, the FIC Act and their equivalents against these two conventions is not academic box-ticking; it is how a domestic prosecutor demonstrates, where extradition or mutual legal assistance to a treaty partner is required, that the domestic predicate satisfies dual criminalityDual criminalityThe requirement, in extradition and mutual legal assistance, that the conduct alleged be a crime in both the requesting and requested states..

Self-launderingSelf-launderingThe criminalisation of the original perpetrator of a predicate offence for subsequently laundering the proceeds of that same offence. is the second concept every new investigator must internalise early, because instinct again misleads. Many people assume that "laundering" necessarily involves a third party; the accountant, the banker, the lawyer who cleans someone else's money. In fact the FATF standard and the great majority of domestic statutes criminalise self-launderingSelf-launderingThe criminalisation of the original perpetrator of a predicate offence for subsequently laundering the proceeds of that same offence.: the original perpetrator of the predicate offencePredicate offenceA specified underlying criminal offence whose proceeds are alleged to be the subject of a laundering charge. who then converts, transfers, conceals or acquires the proceeds of his own crime commits a separate, additional offence of money laundering, cumulative to whatever charge attaches to the predicate itself. This matters practically in three ways. First, it allows a second conviction and often a materially longer aggregate sentence, which is why plea negotiations in complex fraud cases so often turn on whether the laundering count will be pursued alongside the fraud count. Second, it allows prosecution to proceed on the laundering count even where the predicate prosecution fails for a technical or jurisdictional reason, provided the underlying criminal origin of the funds can still be proven to the requisite standard, a point I develop further in the fourth lesson of this module. Third, it creates an incentive structure investigators should exploit deliberately: because self-launderingSelf-launderingThe criminalisation of the original perpetrator of a predicate offence for subsequently laundering the proceeds of that same offence. is typically easier to prove through financial-flow evidence than the underlying predicate is through witness or forensic evidence, a well-sequenced investigation sometimes builds the money trail first and uses it to corroborate, or even to trigger, the predicate investigation, rather than waiting for the predicate case to be complete before touching the financial side.

A minority of jurisdictions, largely in continental Europe though the number has shrunk over successive FATF mutual evaluation rounds, retain some form of "self-launderingSelf-launderingThe criminalisation of the original perpetrator of a predicate offence for subsequently laundering the proceeds of that same offence. exemption" reflecting an older doctrinal view that a person cannot be said to "launder" what was already his by virtue of having stolen it — the ne bis in idem or double-jeopardy intuition. FATF mutual evaluations have consistently flagged the retention of any such exemption as a technical-compliance deficiency against Recommendation 3, and the trend across the 2020s evaluation cycle has been toward its elimination, most visibly in the EU's AMLD6 (Directive (EU) 2018/1673, since absorbed into the broader 2024 AMLR/AMLD6 recast package) which expressly requires member states to ensure self-launderingSelf-launderingThe criminalisation of the original perpetrator of a predicate offence for subsequently laundering the proceeds of that same offence. is punishable. When you assess a jurisdiction's AML framework for adequacy — whether for a mutual evaluation submission, a correspondent-bank due-diligence questionnaire, or a comparative-law essay, checking whether self-launderingSelf-launderingThe criminalisation of the original perpetrator of a predicate offence for subsequently laundering the proceeds of that same offence. is criminalised, and whether the predicate list is all-crimes or schedule-based, are two of the fastest diagnostic questions available to you, and both go directly to whether "dirty money," once identified, can actually be prosecuted as such.

For most of the twentieth century, tax offences sat in a separate legal universe from money laundering. Tax administrations pursued their own assessment, penalty and prosecution machinery; financial intelligence units and prosecutors chasing laundering built their cases around drugs, fraud and organised crime. That separation collapsed, formally, in February 2012, when FATF revised its 40 Recommendations to add "tax crimes (related to direct taxes and indirect taxes)" to the designated categories of predicate offences discussed in the previous lesson. The 2012 revision did not create a single harmonised definition of "serious tax crimeSerious tax crimeA domestically defined category of tax offence, typically threshold-based, that FATF's 2012 revision requires jurisdictions to treat as a laundering predicate.", it deliberately left that definition to domestic law, consistent with the general FATF approach of setting outcomes rather than dictating statutory text — but it obliged every FATF member and every jurisdiction subject to mutual evaluation to ensure that at least serious tax offences are capable of founding a laundering charge. I regard this as one of by some distance the most consequential technical changes to the AML architecture in the last two decades, because it converted the tax administration, previously a bystander to the laundering fight, into a frontline predicate-offence investigator whose case files now matter directly to the FIU and to prosecutors pursuing POCA-style confiscation.

Before you can apply that 2012 change competently, you need the three-way distinction I press on every cohort: avoidance, aggressive avoidance and evasion sit on a single spectrum, but only the far end of that spectrum is criminal. Tax avoidance in its classical, uncontroversial sense is the lawful arrangement of one's affairs to minimise tax liability within the clear intent of the law — claiming an available deduction, choosing a lower-taxed but genuine corporate form, timing a disposal to fall in a more favourable tax year. Aggressive tax avoidanceAggressive tax avoidanceTechnically lawful arrangements exploiting gaps or mismatches against the law's evident purpose; civilly, not criminally, remediated. occupies the contested middle ground: technically lawful structures that exploit gaps, mismatches or the letter of the law against its evident purpose, of the kind targeted by general anti-avoidance rules (GAARs), the OECD's Base Erosion and Profit Shifting (BEPS) Action Plan, and increasingly by mandatory disclosure regimes such as the EU's DAC6. Aggressive avoidance is not, by itself, a predicate offence for laundering, because it lacks the element of criminality; a GAARGAARGeneral anti-avoidance rule empowering a revenue authority to re-characterise or disregard a transaction lacking commercial substance. re-characterisation typically produces a civil tax reassessment and penalty, not a criminal conviction, though repeated and egregious abuse can shade into fraud where deliberate misrepresentation is present. Tax evasion, by contrast, is the unlawful, dishonest non-declaration or under-declaration of income or the unlawful claiming of relief not due, and it is this category, where a jurisdiction defines it as a serious offence, typically by reference to a monetary threshold or a term-of-imprisonment threshold — that the 2012 FATF revision brings within the predicate-offence net.

The practical difficulty officials face daily is that this spectrum is not self-evidently divided in a single transaction; the same offshore structure can sit anywhere along it depending on the taxpayer's state of mind and on whether the underlying facts presented to the tax administration were true. I tell trainees that the diagnostic question is never "was this structure clever" but "was a material fact concealed or misrepresented to the revenue authority, and was that concealment deliberate." Where the answer is yes, you are in evasion territory, and if the jurisdiction's threshold for "serious" tax crime is met, South Africa's Tax Administration Act 28 of 2011 criminalises a range of conduct including the wilful evasion or assisting in the evasion of tax, and, read together with POCA's all-crimes predicate definition, wilful tax evasion above the threshold of "unlawful activity" generating "proceeds of unlawful activities" is unambiguously capable of grounding both a POCA confiscation and a laundering charge under the FIC Act framework. Where the answer is no — the facts were fully and accurately disclosed and the dispute is purely one of legal characterisation — you are in avoidance or aggressive-avoidance territory, and the correct remedy is a civil reassessment, not a criminal referral, however aggressive the underlying planning.

The OECD's institutional response to the 2012 FATF shift has been to develop the practical toolkit tax administrations need to actually detect and refer serious tax crimeSerious tax crimeA domestically defined category of tax offence, typically threshold-based, that FATF's 2012 revision requires jurisdictions to treat as a laundering predicate., most comprehensively expressed in its "Fighting Tax Crime, The Ten Global PrinciplesTen Global PrinciplesThe OECD's maturity-model framework (2017, periodically updated) for effective tax-crime investigation and inter-agency cooperation." framework, first published in 2017 and periodically updated since, most recently with country-specific implementation reports continuing through the 2020s. The Ten Global PrinciplesTen Global PrinciplesThe OECD's maturity-model framework (2017, periodically updated) for effective tax-crime investigation and inter-agency cooperation. set out a maturity model against which a tax administration's criminal tax investigation capability can be benchmarked: having the tax offence itself properly defined in law, having effective powers of access to information (including beneficial-ownership and banking data), having available a range of sanctions proportionate to the offence, identifying tax crimes proactively rather than only reactively, having a dedicated and adequately resourced investigative capability, having effective tools to enable inter-agency cooperation domestically, exchanging information internationally, protecting the confidentiality of information exchanged, ensuring good governance including safeguards against abuse of investigative powers, and providing the necessary resources across the whole system for the framework to actually function. The principles are deliberately structured as a whole system, not a checklist to be partially satisfied, because the OECD's own peer-review work has found repeatedly that jurisdictions with strong legal powers on paper but weak inter-agency cooperation, or strong investigative capability but no confidentiality safeguards for exchanged information, fail to convert the legal architecture into actual prosecutions.

South Africa's own experience illustrates both the promise and the friction of this integration. The Zondo Commission's findings on state captureState captureSystemic manipulation of public institutions and procurement by private interests for illicit gain, documented in South Africa by the Zondo Commission. documented, in extensive and specific detail, how public procurement fraud and corruption at state-owned enterprises generated proceeds that were then laundered through a combination of consulting-fee structures, offshore shell entities and, in some of the most notorious sequences, direct facilitation by international accounting and advisory firms; several of those same fact patterns also involved undeclared income and fraudulent VAT and corporate-tax positions that SARS has since pursued both through civil assessment and, in the more serious instances, through criminal referral working alongside the National Prosecuting Authority and the FIC. The FATF's February 2023 greylisting of South Africa specifically cited weaknesses in the pursuit of complex money-laundering cases involving state captureState captureSystemic manipulation of public institutions and procurement by private interests for illicit gain, documented in South Africa by the Zondo Commission. proceeds and in beneficial-ownership transparency, both of which touch directly on tax-crime predicate work; South Africa's subsequent remediation program, culminating in delisting in October 2025, included measurable improvements to inter-agency case referral between SARS, the FIC and the NPA that map closely onto several of the OECD's Ten Global PrinciplesTen Global PrinciplesThe OECD's maturity-model framework (2017, periodically updated) for effective tax-crime investigation and inter-agency cooperation., particularly the principles on inter-agency cooperation and proactive identification. The lesson for the working official is that tax crime as a predicate offence is not just a doctrinal add-on to the AML framework; it is now one of the primary channels through which grand corruption and state-capture proceeds are actually prosecuted, precisely because the paper trail a tax audit generates; misdeclared income, undisclosed offshore accounts, fictitious invoices, is often more complete and more readily obtained than the evidence needed to prove the underlying corrupt act itself.

Once a predicate offence and its proceeds are identified, the next analytical task is a precise one that investigators frequently blur: distinguishing proceeds from instrumentalities, and tracing tainted property once it has been mixed with legitimate funds. These distinctions are not academic hair-splitting; they determine which forfeiture regime applies, what evidence must be led, and, ultimately, whether a court can lawfully deprive a respondent of an asset.

Proceeds of crimeProceeds of crimeProperty or economic advantage derived directly or indirectly from the commission of an offence. are property, or an economic advantage, derived directly or indirectly from the commission of an offence, the bribe itself, the fraudulently obtained loan, the profit on a market-manipulation trade. Instrumentalities, sometimes called "property used in the commission of" an offence, are a distinct legal category: the vehicle used to transport trafficked goods, the property used as a base for a drug-manufacturing operation, the bank account used as a conduit even where the funds passing through it were not themselves generated by the offence. Most modern confiscation statutes, including POCA's Chapter 5 and Chapter 6 (which I use throughout this lesson as the worked jurisdictional example because South Africa's dual-track forfeiture architecture is unusually well documented and pedagogically clear), treat proceeds and instrumentalities as separately definable categories, each supporting forfeiture on a different factual showing: proceeds forfeiture requires proof that the property represents the benefit derived from unlawful activity, while instrumentalityInstrumentalityProperty used to commit or facilitate an offence, forfeitable regardless of whether it embodies criminal value itself. forfeiture requires proof that the property was used to commit, or to facilitate the commission of, an offence, regardless of whether the property itself has any illicit "value" attributable to a crime.

Tracing is the evidential and doctrinal mechanism by which an investigator or a court follows value from its criminal origin through subsequent transactions to its current resting place. English and Commonwealth equity lawyers draw a classic distinction, inherited into most POCA-style statutes, between following and tracing in the strict sense. Following is the process of tracking the very same asset as it moves from hand to hand; the same banknotes, the same specific shares, the same identified motor vehicle, without any change in its legal character. Tracing in the strict sense is required once the original asset has been exchanged, converted or substituted for a different asset — cash deposited into a bank account and then used to purchase a property, for instance — and the law permits the claimant (or the state, in a confiscation context) to identify a proportionate claim over the substitute asset because it represents the traceable proceeds of the original property, even though it is a physically different thing. Both following and tracing depend on maintaining an unbroken evidential chain of substitutions; where that chain cannot be reconstructed, the claim to the substitute asset fails, however strong the suspicion.

The single hardest practical problem in this area is the mixed fund: a bank account, business, or investment portfolio into which both licit and illicit money has been deposited or invested, such that no physical segregation of "dirty" and "clean" dollars is possible. Courts across common-law jurisdictions have developed several competing approaches to allocate a claim against a mixed fund. The "lowest intermediate balance" rule, drawn from English trust law, limits the traceable claim to the lowest balance the account held at any point between the tainted deposit and the date of the claim, on the theory that funds cannot be traced into money that was never actually present in the account at the relevant time. Pro-rata apportionment approaches instead treat the tainted and untainted contributions as sharing proportionately in the fund and in any subsequent gains or losses, which is the approach more commonly adopted in statutory civil-forfeiture regimes because it avoids the sometimes arbitrary results the lowest-intermediate-balance rule can produce where large legitimate deposits and withdrawals occur after the taint. South African courts applying POCA have generally favoured a pragmatic, fact-specific approach to mixed funds rather than rigidly adopting either English doctrine wholesale, asking essentially what a fair and evidentially supportable allocation is on the facts presented, which places a premium on the forensic accountant's reconstruction of account activity rather than on doctrinal purity.

POCA's own structure is instructive precisely because it embodies the civil/criminal forfeiture distinction that every jurisdiction wrestling with this problem must eventually resolve. Chapter 5 of POCA establishes confiscation orders, a conviction-based mechanism: following a criminal conviction, the court may make a confiscation order requiring the convicted person to pay an amount equal to the value of his or her proceeds of unlawful activities, calculated according to statutory presumptions that shift the evidential burden onto the convicted person to show that particular property or receipts were not derived from crime once the state has shown a pattern of criminal lifestyle or benefit. Chapter 6, by contrast, establishes civil forfeiture, proceeding entirely independently of any criminal prosecution or conviction, against the property itself (an in rem proceeding) rather than against a person, on the civil standard of proof (balance of probabilities) rather than the criminal standard (beyond reasonable doubt). Civil forfeiture under Chapter 6 requires the state to obtain a preservation order freezing the property, followed by a forfeiture order establishing that the property constitutes either the proceeds of unlawful activity or an instrumentalityInstrumentalityProperty used to commit or facilitate an offence, forfeitable regardless of whether it embodies criminal value itself. of an offence listed in POCA's schedule, and it is available even where no natural person has ever been charged, which is precisely why it has become the preferred tool in cases where the human perpetrator has died, fled the jurisdiction, or cannot be identified with the certainty a criminal charge requires.

The civil/criminal forfeiture distinction is not a peculiarly South African invention; it echoes the United States' long-standing civil asset forfeiture regime under 18 U.S.C. § 981 and related statutes, and the United Kingdom's Proceeds of CrimeProceeds of crimeProperty or economic advantage derived directly or indirectly from the commission of an offence. Act 2002, which similarly maintains parallel conviction-based confiscation and non-conviction-based civil recovery tracks. What South Africa's POCA usefully demonstrates for a comparative teaching purpose is the deliberate design choice to keep the two tracks procedurally and doctrinally separate rather than blending them, precisely because the lower civil standard of proof is constitutionally defensible only where the proceeding targets property rather than liberty, and any blurring of that line invites a due-process challenge. I tell investigators building a forfeiture case that the single most important early decision is which track to pursue, because the evidence-gathering strategy differs from day one: a Chapter 5 confiscation strategy requires you to build toward a conviction first, with asset tracing largely following the criminal case; a Chapter 6 civil forfeiture strategy allows asset tracing and preservation to proceed in parallel with, or even entirely independently of, any criminal investigation, and is frequently faster precisely because it does not have to satisfy the criminal standard of proof or wait for a criminal trial to conclude.

The final, and in casework the most consistently underestimated, step in the proceeds-of-crime chain is the evidential bridge from a predicate offence to a sustainable laundering conviction or forfeiture order. Investigators frequently assume that once a predicate offence is plausible and a suspicious pattern of financial activity is documented, the laundering case follows naturally. It does not. Every laundering charge requires the prosecution to establish, to whatever standard of proof applies, that the property in question is in fact the proceeds of unlawful activity, and this "unlawful origin" element is very often the most contested issue at trial, more so than the identity of the person who moved the money.

There are two broad evidential routes to proving unlawful origin, and a competent prosecutor chooses between them, or combines them, deliberately rather than by default. The first is direct proof of the predicate offence: securing a conviction, or at minimum leading admissible evidence, that a specific identified crime occurred and that the property in question is traceable to it. This is the doctrinally cleanest route and the one instinctively preferred by prosecutors trained in conventional criminal procedure, but it is frequently unavailable in practice — the predicate may have occurred abroad and be unprosecutable domestically, the predicate offender may be dead, unidentified, or have fled beyond extradition reach, or the predicate case itself may simply be too weak on its own evidence to sustain a conviction even though the financial pattern surrounding it is compelling.

The second route, and the one that has become steadily more important across common-law and increasingly civil-law jurisdictions, is inference-based proof of unlawful originInference-based proof of unlawful originEstablishing that property is proceeds of crime through a constellation of circumstantial markers, without proving a specific predicate offence.: establishing, through the pattern, scale, structuring and surrounding circumstances of the financial activity itself, that the property could only realistically be explained as proceeds of crime, without necessarily proving which specific predicate offence generated it. Courts have developed several recognised inferential markers that, cumulatively, support such a finding: income or asset accumulation grossly disproportionate to any lawful, declared source of income; the use of concealment techniques (shell companies, nominees, structuring, cash-intensive fronts) that have no plausible legitimate commercial rationale; false or contradictory explanations offered by the respondent when asked to account for the property's source; association with persons independently known to be engaged in criminal enterprise, and a pattern of transactions matching a known laundering typology (rapid layering through multiple jurisdictions, round-tripping, back-to-back loan structures) rather than any ordinary commercial logic. No single marker is sufficient alone, but English, South African, Australian and increasingly EU jurisprudence has accepted that a sufficiently strong constellation of these markers can discharge the prosecution's burden even absent a specific proven predicate, particularly in civil forfeiture proceedings where the standard of proof is lower.

This inferential approach is not an evidentiary shortcut invented to make prosecutors' lives easier; it reflects a considered policy response to the practical reality that sophisticated launderers deliberately engineer their affairs so that the specific predicate offence is undiscoverable or unprovable, while the pattern of concealment remains, almost definitionally, harder to disguise completely. FATF's own guidance and the Palermo Convention's traveaux préparatoires both recognise that requiring proof of a specific predicate offence in every case would render the laundering offence largely ineffective against professional, organised laundering networks, and most jurisdictions' laundering statutes are drafted (through phrases such as "reasonable grounds to believe" or "ought reasonably to have known") precisely to accommodate inference rather than demand certainty as to the specific predicate.

Standards of proof, however, diverge sharply across jurisdictions and across the criminal/civil forfeiture distinction examined in the previous lesson, and this divergence is probably the most important practical variable an investigator must track when deciding how to structure a case, particularly one with a cross-border dimension. In a criminal laundering prosecution, the prosecution must prove unlawful origin beyond reasonable doubtBeyond reasonable doubtThe criminal standard of proof requiring the fact-finder to be satisfied to a near-certain degree, applicable to criminal laundering convictions. (or its civil-law equivalent, intime conviction or a comparably demanding threshold), and courts in most jurisdictions require the inferential markers described above to be strong enough to exclude any reasonable innocent explanation, not simply to make the criminal explanation the more probable one. In a civil, non-conviction-based forfeiture proceeding of the kind POCA Chapter 6 authorises, the applicable standard drops to the balance of probabilitiesBalance of probabilitiesThe civil standard of proof requiring only that a proposition be more likely true than not, applicable to non-conviction-based forfeiture., meaning the state need only show that it is more likely than not that the property represents proceeds of unlawful activity, a materially lower bar that explains why civil forfeiture has become the preferred vehicle in exactly the cases where the inferential evidence is strong but a specific predicate cannot be proven to the criminal standard. A third standard, intermediate in character, arises in some administrative and regulatory contexts — a supervisory authority imposing an administrative sanction for AML control failures typically applies a standard closer to civil balance-of-probabilities, even where the underlying suspected conduct, had it been prosecuted criminally, would have required proof beyond reasonable doubtBeyond reasonable doubtThe criminal standard of proof requiring the fact-finder to be satisfied to a near-certain degree, applicable to criminal laundering convictions..

The practical discipline I try to instil in investigators building a case file is to document the evidential route chosen from the outset, rather than allowing the case to drift between direct and inferential proof without a clear analytical structure. A file that pursues direct proof of predicateDirect proof of predicateEstablishing a specific identified predicate offence and tracing property to it, typically via a prior or parallel conviction. should be built around the elements of that specific predicate offence, the corrupt payment, the fraudulent invoice, the trafficked shipment; with the financial-flow evidence used corroboratively. A file that pursues inference-based proof should instead be built systematically around the recognised markers: a full asset-and-income reconciliation demonstrating disproportion, a documented account of every concealment technique identified, a complete record of any false explanation offered by the respondent (ideally obtained under a properly administered interview or affidavit process so that the falsity itself becomes usable evidence), and an expert typology analysis situating the observed pattern against known laundering methodologies. Where a case can plausibly support both routes, pleading them in the alternative, or pursuing an inference-based civil forfeiture in parallel with a slower-moving criminal referral, is frequently the most resilient strategy, because it protects the asset through preservation while the criminal case, if pursued, is built on its own timeline. What I want every official finishing this module to carry forward is that "proving the money is dirty" is not a single evidentiary event but a structured choice between distinct evidential architectures, each with its own standard of proof, its own required record-keeping discipline, and its own vulnerability on appeal if the wrong architecture is chosen for the facts actually available.