Two questions, routinely merged
Source of funds asks where the specific money used in this transaction came from. The answer is a particular payment: this salary payment, this property sale, this transfer from that account.
Source of wealth asks how the customer's overall wealth was accumulated. The answer is a narrative: a career in a given profession, a business sold in a given year, an inheritance, an investment portfolio built over time.
They answer different questions and a customer can satisfy one while failing the other. A customer who deposits from a bank account in their own name has a clear source of funds; if their declared occupation could not plausibly generate the amounts involved, their source of wealth is unexplained. Conversely a genuinely wealthy customer whose deposits arrive from an account belonging to somebody else has a clear source of wealth and a source of funds problem.
Most frameworks require both for enhanced due diligence cases. Most operators collect one and label it with both names.
Why this is harder in gambling than in banking
A bank sees a customer's balance, their income arriving, their outgoings, and typically a long relationship history. A gambling operator sees deposits from an instrument and nothing else.
That asymmetry has three consequences.
The deposit tells you almost nothing. It tells you money moved from an instrument bearing a name. It does not tell you whose money it was, whether it was earned, borrowed, gifted, taken from a joint account or provided by a third party who controls the customer.
The plausible-sounding explanation is unusually available. Winnings explain money. A customer who has genuinely won elsewhere, or who can produce a withdrawal record, has a documented explanation that is hard to disprove and that also explains an absence of tax records in jurisdictions where winnings are untaxed. Treating a claimed prior win as evidence without corroboration is one of the more common weaknesses in this area.
The commercial pressure runs the wrong way. The customers who trigger source of funds enquiries are the highest-value ones. The enquiry is intrusive, it frequently ends the relationship, and the person asking is usually in a smaller team than the person whose target it affects. Every structural safeguard in the governance lesson exists because of that pressure.
Evidence, and what each kind actually proves
Bank statements are the most informative single document. They show income arriving, outgoings, other gambling activity and the state of the account. They are also the most intrusive, and they are the document most often edited. Obtain them from the customer through a channel that makes alteration harder where possible, and read them rather than filing them.
Open banking is bank statements without the alteration risk, delivered through a consented direct connection. It is the strongest routinely available source and the only one that reveals gambling spend with other operators. Its constraint is consent, and a customer declining consent is information of a kind, though not conclusive.
Payslips and employment evidence establish regular income. Verify the employer exists and that the figures are internally consistent, because fabricated payslips are cheap and common. A payslip for an amount that cannot support the observed deposits explains nothing, and filing it as though it did is the error to avoid.
Tax documents are among the strongest evidence because fabricating them carries additional legal exposure, and because they are usually produced by or filed with an authority.
Sale and transaction records for property, a business, a vehicle or an investment establish a one-off inflow. Check that the sum and the date correspond to the deposits in question.
Inheritance documentation such as grants of probate or solicitor correspondence establishes a legacy. The common gap is the delay between the event and the deposits: a substantial inheritance three years before the deposits began needs an explanation of where the money sat in between.
Prior gambling winnings require corroboration from the paying operator, not the customer's account of them. A withdrawal record from a licensed operator is evidence; a screenshot is not.
Crypto provenance requires chain analytics to attribute the source address and assess exposure to sanctioned or illicit counterparties, plus an explanation of how the customer acquired the assets. A transfer from a regulated exchange carries originator information under travel rule implementations; a transfer from a self-custodied wallet carries none, and the burden falls entirely on the customer's explanation and the analytics.
Assessing rather than collecting
The test is not whether a document was obtained. It is whether the evidence, taken together, explains the activity.
Four questions structure the assessment.
Does the amount work? Compare the declared income or wealth against the observed deposits over a period. A customer depositing more per month than their declared annual income can support has not been explained, whatever paperwork is on file.
Does the timing work? A one-off inflow should correspond in time to the activity it is said to fund. Large gaps need their own explanation.
Is it internally consistent? The occupation, the location, the age, the payment instruments and the stated wealth should cohere. Incoherence is the most reliable signal available and requires no external data.
Is it corroborated? Independent evidence beats the customer's assertion. Where only the assertion exists, the file should say so, because a later reader needs to know which parts were verified.
The failure mode to avoid is the file that contains documents and no conclusion. A supervisor reading it will ask what the operator concluded and why, and "we obtained source of funds evidence" is not a conclusion.
When the answer is unsatisfactory
This is the point at which the framework is real or is not.
If the evidence does not explain the activity, the operator has a decision to make and a very limited set of acceptable options. Obtain more evidence. Restrict the relationship. Terminate it. In each case, consider whether the circumstances give rise to suspicion, because if they do, the reporting obligation arises independently and immediately.
What is not available is continuing the relationship unchanged on the basis that the customer was asked. An operator that requested source of funds, received an inadequate answer, and carried on is in a worse position than one that never asked, because it has documented that it knew the activity was unexplained.
Two constraints shape how this is handled with the customer.
Tipping off. If a report has been made or is contemplated, the customer cannot be told in terms that might prejudice an investigation. This is why financial crime restrictions are communicated tersely and without reasons, and why staff need a script that does not improvise an explanation.
Returning funds. Where an operator holds funds it suspects are criminal property, returning them to the customer may itself be an offence in some regimes. The sequence, whether consent or a defence is required before dealing with the funds, is jurisdiction-specific and is exactly the situation in which the nominated officer takes the decision rather than an account manager.
Proportionality without a loophole
Not every customer warrants this. A framework that demands documentary source of wealth from ordinary recreational customers is disproportionate, expensive and will be circumvented internally.
The workable structure is the tiered one used across the sector: no enquiry for the great majority, frictionless background checks in the middle, customer-facing enquiry at higher levels or on a risk trigger, and documented evidence required before continuing at the top or where earlier tiers raised something.
Two things stop the tiering becoming a loophole. Thresholds must be set with recorded reasoning, before the commercial consequences are known, by someone who does not carry the revenue target. And any risk trigger must be able to pull a customer up the tiers regardless of their spend level, because the whole point is that a modest-looking account can be the problem.