From Fragments to a Shared Picture: Rethinking the Fight Against Fraud

18 November 2026

Fraud is often described as a data problem. That is only partly true: most countries already possess vast quantities of fraud data. The deeper problem is that the information is fragmented. A bank sees the payment. A telecommunications company sees the call. A digital platform sees the advert or account. The police see the crime report. And the victim experiences the whole journey.

The criminal is often the only participant who sees the entire operation from beginning to end.

If criminals organise across sectors, our response must organise across sectors. Information sharing is not an administrative support to fraud policing; it is part of the operational response. Designed properly, it can do more than identify offenders: it can prevent harm, protect repeat victims, preserve funds and return money to the people and organisations from whom it was stolen.

 

A trusted network, not simply a database

Cifas was established in 1988; today our community includes more than 800 organisations across 25 sectors, from banking, lending and insurance to telecommunications, retail, technology and the public sector.

Our operating idea is reciprocity: members contribute confirmed fraud-risk information and, in return, gain access to information contributed by others. Fraud prevention is a shared capability, not a competitive advantage guarded by one institution.

In 2025, members filed more than 440,000 cases to our National Fraud Database and prevented more than £2.4 billion in fraud losses. The National Fraud Database holds approximately two million records, searchable 24 hours a day.

But scale alone does not create trust; trust comes from rules, assurance and accountability. Cifas is a Specified Anti-Fraud Organisation under section 68 of the Serious Crime Act 2007, giving public authorities a legal gateway to share information with us, and members work within detailed handbooks governing purpose limitation, transparency, lawfulness, data quality, proportionality and security.

A database match is not, by itself, proof of fraud. Before filing a case, a member must have reasonable grounds to believe that fraud or financial crime has occurred or been attempted, supported by clear, relevant and sufficient evidence, and the individual's explanation must be considered. Members cannot simply reject an application because a name appears in the database.

That distinction matters. Effective information sharing requires two things at once: the confidence to share and the discipline not to overreach.

 

Four complementary forms of sharing

Our model has developed beyond a single database. The National Fraud Database shares confirmed external fraud risk, including identity fraud, facility takeover, false applications and money-mule activity. The Insider Threat Database allows members to share cases of dishonest or seriously improper conduct by staff, applicants and contractors.

Our Scam Victim Check service serves a different purpose: it lets participating organisations identify people already confirmed as scam victims, so that another provider can recognise heightened vulnerability and apply safeguards. It protects people at risk of being targeted again; it is not a database of offenders.

ScamLink addresses the infrastructure of scams. Members share signals such as malicious URLs, domains, email addresses, telephone numbers, social-media handles and IP addresses, routed to partners able to investigate or disrupt them. This tackles a fundamental weakness: the organisation that discovers a scam is often not the organisation capable of taking it down.

These services have different purposes and operate at different evidential thresholds: confirmed case data supports prevention and investigation, intelligence points investigators towards emerging threats, victim data enables safeguarding, and scam signals support disruption. Treating all four as the same would create excessive risk or make the system too cautious to be useful.

 

Connecting the private sector to policing

Every day, Cifas data is supplied into the UK policing system. The City of London Police, the national lead force for fraud, receives reports from the public and businesses, and its Report Fraud Analysis Services collates fraud and cybercrime intelligence into investigation packages for police forces.

The practical value is that private-sector investigations do not remain trapped inside individual companies: a suspicious application at a lender, an account takeover at a telecoms company and a mule account at a bank may be parts of the same network, connected by shared identifiers.

Data transfer alone is not partnership. It also requires shared priorities, common definitions, feedback loops and people who can translate between commercial operations and policing.

 

From isolated incidents to an organised network

A recent Cifas operation, Operation Carrow, shows what coordinated intelligence can achieve. In mobile dealer fraud, a fraudster contacts a genuine mobile customer, impersonating a network operator or retailer and offering an attractive upgrade, then uses social engineering to take control of the account or place an order.

A high-value handset is delivered to the real customer; the fraudster claims the wrong device was sent and instructs the customer to return it, in reality to a mule address. The handset is then resold online, moved through complicit shops or exported.

Seen by one company, this looks like a customer-service incident; seen across multiple companies, delivery routes and addresses, it becomes organised criminality.

We brought together telecommunications providers, the City of London Police, regional policing, our postal operator and other partners. The intelligence cell established common reporting, identified repeat addresses and connected entities, turning separate company records into a national picture. By 16 September 2026, mobile network operators had recorded more than 15,000 incidents during the year, an average of 58 a day and an increase of 85 per cent compared with the same period in 2025. Coordinated work had helped identify more than 700 mule addresses, at least 300 of which had been active within the preceding 12 months.

The lesson is not limited to telecommunications. A repeated address, telephone number, device, beneficiary account or digital identifier may look unimportant inside one institution. Across a network it can become the organising feature of an investigation.

 

Reimbursement is an important protection, but it is not recovery

The UK has changed how victims of authorised push-payment scams are protected: mandatory reimbursement has applied to eligible APP scam payments since 7 October 2024, covering individuals, microenterprises and charities. Sending and receiving payment firms share the cost, giving both sides an incentive to prevent the fraud.

In the 18 months to 31 March 2026, 88 per cent of the money lost in in-scope claims, £316 million, was reimbursed; 82 per cent of claims were closed within five business days and 98 per cent within 35.

But reimbursement is not recovery: it restores the victim's financial position by transferring the loss to payment providers. Unless criminal funds are traced and recovered, the offender may still retain the proceeds.

The regime also has limits: it applies to specified UK account-to-account payments, not to every victim, payment method or overseas transfer, and cannot be the whole response. The strategic objective should be prevention first, rapid reimbursement where appropriate, and recovery wherever funds can still be traced.

 

Learning the lessons

Many nations are starting to adopt this approach. Korea is moving strongly in this direction. Its Anti-phishing Sharing and Analysis Platform (ASAP) enables real-time sharing of vishing-related data across the financial sector, and recent reforms have expanded information sharing across finance, telecommunications, investigative agencies, secondary financial institutions and virtual-asset exchanges. Similar developments are under way elsewhere.

The UK experience suggests five design principles to shape what happens next:

  1. Build reciprocity into the operating model. Organisations that benefit from pooled intelligence should also contribute relevant data.
  2. Separate confirmed case data, intelligence, victim-safeguarding data and disruption signals. 
  3. Establish a neutral and trusted coordinating body. No single bank, telecoms company, platform or police unit sees the entire threat.
  4. Create operational feedback. Contributors need to know whether shared intelligence produced prevention, disruption or recovery; feedback improves both participation and data quality.
  5. Make victim recovery an investigative objective from the first hour, not an administrative task after the criminal case.

A bank sees the payment. A telecommunications company sees the call. A platform sees the account. A delivery company sees the parcel. The police see the report. The victim experiences the whole crime.

Our task is to make sure the criminal is no longer the only person who can see the entire operation.

Trusted information sharing converts thousands of partial views into one actionable picture. Public-private partnership lets the organisation that spots the threat reach the organisation that can stop it. A victim-centred recovery strategy moves us beyond recording loss, towards preserving funds and returning them.

The measure of success should not be how many reports we collect, databases we build or offenders we identify. It should be how much harm we prevent, how quickly we disrupt criminal infrastructure, and how much stolen money we return to victims. That is the opportunity and challenge for all of us.

Share:
Posted by: Mike Haley

Mike is Chief Executive of Cifas.

NEXT POST
 

The evolution of AI-generated documents

7 October 2026

AI-generated documents are rapidly becoming more convincing and accessible, enabling fraudsters to create or alter financial and identity documents with minimal skill. Organisations must move beyond visual checks and adopt stronger detection, verification and cross-application analysis techniques to identify document fraud.

CONTINUE READING
Back to blog home >
Posted by: Mike Haley

Mike is Chief Executive of Cifas.

Categories