FIGHTING Fraud © Ikon Images / Alamy
Total Page：16
File Type：pdf, Size：1020Kb
This is an independent publication by Raconteur Media _ 21. March. 2011 FIGHTING FrAUD © Ikon Images / Alamy whether it is accounts payable data payments over £10,000, so the clerk BeatinG FraudSTERS at THEIR own GAME or voter registrations, Benford’s Law, ensures all the payments to the sus- also known as the first-digit law, pre- pect supplier are below that level. BUSINESS RESILIENCE Fraud takes many forms and the digital economy dicts that real-world data get distrib- That, by itself, is not necessarily uted in a certain non-uniform way. If suspicious. However, forensic anal- has created many new opportunities for fraudsters. But technology is also the suspect data deviate significantly ysis reveals a large number of pay- helping businesses fight back against fraud, which costs the UK economy a from the expected distribution, there ments have 7, 8 or 9 as their first digit, Geoff Nairn could be a fraudster at work. eg, £9,450 or £708. Benford’s law pre- staggering £38 billion a year. reports In the business world, frauds typi- dicts, counter-intuitively, that the cally centre on procurement, payroll higher numbers should each occur as and expenses. So, when a business the first digit around 5 per cent of the Ȗ While information technology has importance. “In the past, we were the accounting records and other types suspects fraud, forensic account- time, while the number 1 should be made our lives easier, it also opens poor relation in the accounting pro- of ‘structured’ data to find unusual ants start by analysing the accounts the first digit in 30 per cent of cases. new doors for tech-savvy criminals fession, but forensic data analytics patterns that betray many types of payable data. A quick check then reveals that the to commit fraud. Fortunately, tech- is now a hot area,” says Paul Walker, corporate fraud. Benford’s Law is not the only address given for the suspect sup- nology can also be used to detect head of forensic technology and dis- weapon in their armoury, of course, plier corresponds to a domestic resi- fraud by analysing data and search- covery services at Ernst & Young. DATA but it can be surprisingly effective. dence rather than a business. The ing for patterns of behaviour that His department has grown from The key to detecting fraud is data – For example, a dishonest clerk has alarm bell rings. betray the fraudsters. just six people three years ago to the more, the better – and a 70-year- set up a fictitious supplier on the sup- As well as structured data, recent In recent years, the specialist field more than 70 today. Traditionally, old mathematical principle called plier database. Company procedure of forensic accounting has grown in the work involves trawling through Benford’s Law. Irrespective of requires additional authorisation for Continued on PAGE 03 TWITTER.COM/RACONTEURMEDIA RACONTEUR 01 FIGHTING FRAUD Distributed in CONTRIBUTORS CLAIRE MANUEL SIMON MICHELL GEOFF NAIRN Former editorial director of Newsdesk A journalist and former business analyst A freelance journalist specialising in IT, Publisher Editor Design Media Group, she is a freelance business and managing editor at the Jane’s Infor- telecommunications and, most recently, Heather Sophia Athié Claire Manuel The Surgery writer and editor mation Group cleantech and sustainability issues For more information about Raconteur Media publications in PETER ARCHER IAN WELSH ROBERT SCHIFREEN The Times and The Sunday Times, please contact Freddie Ossberg Specialist writer on finance and business, A specialist on responsible business and A former computer hacker, Robert is T: 020 7033 2100, E: [email protected], he was formerly with the Press Associa- sustainability issues. He is editor of Ethi- a well-known writer, broadcaster and W: www.raconteurmedia.co.uk tion and NBC cal Corporation magazine trainer specialising in IT security issues The information contained in this publication has been obtained from sources the proprietors believe to be correct. However, ROBERT FOX MICHAEL DEMPSEY Guy Clapperton no legal liability can be accepted for any errors. No part of this A journalist and broadcaster, he is defence Journalist and media trainer who has Author, broadcaster, speaker and journal- publication may be reproduced withoutthe prior consent of the correspondent for the Evening Standard worked for the BBC, the Financial Times ist. His latest book, This Is Social Media, Publisher. © RACONTEUR MEDIA and numerous publications is about to go into its second edition “Today, there is a lot more remote ANALYTICS interaction in financial services so The fraud landscape is constantly it is very difficult to know your cus- changing. The recession has seen tomer,” says Brian Kinch, who was the growth of first-party fraud, in head of customer account fraud at which customers with a good credit LloydsTSB before joining FICO. history suddenly decide to stop pay- This problem particularly affects ing off their debts. This behaviour is the insurance sector, with the growth difficult to detect until it’s too late, of online operators and aggrega- but Mr Kinch says it is an area where tor websites that let people ‘plug data analytics can help. and play’ with their personal data “Analytics shows us that there to obtain a better quote. “Insur- are certain characteristics that are ance fraud has always been there of common to these customers,” he course, but it has been getting pro- explains. For example, they will take gressively worse,” says Mr Ash of SAS. all the products on offer when they When the credit crunch started to open the account and will try to accu- bite a couple of years ago, the era of mulate as much credit as possible easy credit came to an end. Instead of before disappearing. trying to create new accounts, fraud- New types of fraud are emerg- sters switched to hijacking existing ing all the time. One of the most ones using cashpoint scams and unusual insider frauds involved a online phishing attacks. Sainsbury’s IT manager who stole USING TECHNOLOGY TO BEAT FRAUD As banks got wise to these attacks, more than 17 million Nectar points, the fraudsters switched to identity worth £80,000, using a series of fake fraud as a way to obtain products and accounts. He was jailed earlier this Continued From PAGE 01 historical transactional data of a channels and so protect against Data analysis is a services that were not subject to as month. The explosion in social cardholder down to a few variables multi-channel attacks. vital weapon in the strict an approval process as credit media is also attracting fraudsters, years have seen a tremendous that are then used to analyse every “Sometimes there are subtler battle against fraud applications. both as a new channel to deliver growth in the amount of unstruc- transaction in a few milliseconds. threats that, when viewed separately, Identity fraud used to be relatively scams and as a rich source of per- tured data stored in organisations, Brian Kinch, FICO senior part- can appear benign. But when you minor in the UK, but it now accounts sonal information. particularly in regulated industries ner, says the software in effect learns bring them together, you can spot for almost half of all frauds and costs The ingenuity of fraudsters knows where e-mails and telephone con- about the cardholder behaviour in fraud earlier,” says Derek Wylde, the UK economy more than £2.7 bil- no bounds and modern technology versations are now preserved as a a bid to reduce the number of false head of group fraud risk at HSBC. lion a year, according to the National has made their task easier. But the matter of course. positives – legitimate transactions He gives the example of a custom- Fraud Authority. good news is that technology can Unstructured data pose particular flagged as possibly fraudulent. er’s credit card, which is used shortly “Identity fraud is the modern-day also be harnessed as an ally in the challenges for forensic data analysts. For example, if a cardholder fre- after their debit card and followed equivalent of Dickensian pickpock- battle against fraud. Mr Walker admits that searching quents casinos and often makes sev- by activity on the internet banking ets,” says Richard Hurley, communi- though archived phone conversations eral withdrawals of cash from casino channel. Viewed in isolation, the cations manager for CIFAS, the UK’s or e-mail records to find evidence of ATMs, FICO’s software learns this activities are not suspicious, but the fraud prevention service. fraud can be like trying to find a nee- behaviour and knows not to chal- three activities happening within a dle in the proverbial haystack. lenge the withdrawals. For card- short timeframe is suspect. Nevertheless, it is often the only holders who don’t visit casinos, such way to uncover the more sophisti- behaviour is highly suspicious. cated types of fraud and other ‘non- One of the biggest problems facing compliant’ behaviour, such as price- financial institutions when it comes The key to detecting FIGHTING Fraud toGETHER fixing, blackmail or bribery. to fighting fraud is that of information “Our software is able to look for silos. At most banks, the fraud man- fraud is data – the more, Fraud can happen to any organisa- rigorous verification. Know your cus- certain words in a dialogue that indi- agement systems have developed in tion. No-one is exempt and every or- tomer, know the application and cate non-compliant behaviour,” says isolation, with one system for moni- the better ganisation is vulnerable, both to ex- check the application and the sup- Mr Walker. toring credit cards, one for debit cards, ternal attacks from customers intent porting identity documents.