At the recent International Anti-Financial Crime Summit 2026, I joined the panel on AI, Data and the Future of Financial Crime Compliance.
One of the questions I raised was how much financial crime can an individual organisation realistically identify when it is relying only on its own data?
This question is important because criminals do not organise their activity around the structures of the firms trying to stop them. They move between accounts, cards, merchants, payment providers, digital wallets and jurisdictions. Each organisation may therefore see only one small and apparently legitimate part of a much wider chain.
A card transaction viewed in isolation may look perfectly normal. It could be a purchase, refund, cash withdrawal, wallet top-up or transfer.
The risk may only become clear when that transaction is considered alongside activity taking place across other cards, merchants, channels or geographies.
This is increasingly important as fraud and money laundering continue to converge. The proceeds of scams, account takeover, synthetic identities and other fraud can enter the card ecosystem before being moved onwards through merchants, payment providers, wallets or other financial institutions.
The individual transaction may not provide the strongest signal. The relationships between transactions often tell the more important story.
Criminals understand this. By spreading activity across different organisations, they can ensure that every participant sees only a small and plausible fragment of the overall pattern.
Banks and payment firms face considerable regulatory and public pressure to prevent financial crime, protect customers and explain why suspicious activity was not identified sooner.
However, the criminal journey often starts much further up the chain.
A significant amount of activity is likely to have taken place before it reaches the financial institution expected to identify and stop it. A false identity may have been created using AI, a victim contacted through email or social media, or a fake advert posted online. By the time the payment reaches the financial system, the fraud will already be well advanced.
Financial institutions must continue to improve their controls, but there is a wider question about accountability.
Should the organisations providing the identities, communications, technology and access used earlier in the journey carry greater responsibility? Are we sharing enough intelligence between the different organisations that can see each stage of the activity?
Fraud and money laundering do not take place within organisational boundaries. Our approach to preventing them cannot continue to do so either.
The answer is not simply to provide financial crime teams with more data.
Most teams are already managing significant alert volumes, increasing regulatory expectations and pressure on budgets and resources. Adding another source of information without improving decision-making could make the problem worse.
The opportunity is to turn wider network data into useful intelligence that helps firms identify patterns they could not see alone and prioritise the activity most worthy of investigation.
This is where network-level data and AI will make a meaningful difference.
Rather than assessing every transaction in isolation, models can consider activity across a much wider dataset, identify connections and behaviours associated with known money laundering typologies and provide investigators with a clearer indication of where to focus.
The objective is not more alerts. It is be better-quality intelligence, stronger prioritisation and more effective investigations.
Mastercard’s AML Account Risk capability provides a proven example of how this can work.
The solution continuously assesses approved Mastercard transactions processed across the Mastercard network, covering Mastercard-branded credit, debit and prepaid products.
Using machine learning across multiple money laundering typologies, the activity is assessed and aggregated at primary account number level. Higher-risk accounts exceeding the relevant thresholds are then presented to the issuer with a risk score, reason codes and supporting transaction information.
This gives investigators a clearer starting point. They can review the accounts presenting potentially higher risk, understand the activity contributing to the score and decide whether further investigation is warranted.
And just to be clear, this capability is not intended to replace an issuer’s existing transaction monitoring, customer knowledge or investigative processes. It provides an additional network-level view that can complement those controls and identify activity that will not be visible within the issuer’s own environment.
Notifications can be accessed through a dashboard or integrated into an existing AML platform through an API, allowing firms to incorporate the intelligence into their established processes.
Importantly, it remains the issuer’s responsibility to investigate the activity and determine whether money laundering has taken place. This is decision support, not automated proof of criminality so importantly firms retain complete control.
Access to network-level intelligence does not automatically deliver better outcomes.
Firms still need to decide how the information fits into their financial crime framework. They must establish ownership between fraud and AML teams, agree how alerts will be triaged, understand how the data affects existing risk assessments and measure how it improves investigative outcomes.
Fairness, privacy, explainability and governance must also remain central. Financial crime teams need to understand what is driving a risk score and be able to explain how intelligence is being used to support decisions.
The technology can identify an important signal. The operating model determines whether the organisation can act on it effectively.
It is clear from IAFCS2026 that individual organisations cannot be expected to identify increasingly connected financial crime while remaining restricted to their own data.
There will always be information that sits outside the view of an issuer. Network-level intelligence offers an opportunity to bring those disconnected signals together and provide firms with a more complete picture.
Brighter Consultancy is working in partnership with Mastercard to help firms understand how AML Account Risk could complement their existing controls and what changes may be needed to turn the intelligence into measurable improvements.
If you would like to discuss the capability or explore what network-level intelligence could mean for your organisation, we would welcome a conversation.