‘Gone in 60 Seconds’ – and the lessons learned
During late October 2012, fourteen individuals were charged following a Federal Bureau of Investigation (FBI)-led investigation into the theft of over $1 million from Citibank using cash advance kiosks at casinos located in Southern California and Nevada. The fraudsters stole the money by exploiting a gap in the Citibank’s ATM applications—which required multiple withdrawals all within 60 seconds—giving the fraud a popular name among the legal and security circles ‘Gone in 60 Seconds’. 
The modus operandi worked as follows: the main accused recruited conspirators who were willing to open multiple Citibank checking accounts. He then supplied his co-conspirators with “seed” money, which was deposited into the recently opened accounts. After the money was deposited into the checking accounts, the conspirators would travel to nearly a dozen casinos in California, Las Vegas and Laughlin. When inside the casino, the conspirators, used cash advance kiosks at casinos to withdraw (all within 60 seconds) several times the amount of money deposited into the accounts, by exploiting the Citibank’s ‘security gap’ they discovered. The accused were also careful to keep both their deposits and withdrawals under $10,000 in order to avoid federal transaction reporting requirements and conceal their fraud.
What was the ‘security gap’ which could be discovered by the fraudsters and NOT by the Citibank’s IS Auditors?
As long as all of the withdrawals were made within one minute of the first, Citibank’s software assumed the transactions indicated erroneous duplicate processing of the first request, and hence no red flags would be raised. While the sophisticated plot allowed the group to collect more than a $1 million over an eight-month period, thanks to a mundane flaw in the criminals’ logic eventually led the FBI to its suspects: they all used their real names when activating the bank accounts that ended up excessively overdrawn.
What Citibank could have done to prevent this?
Presumably, the control weakness escaped the multiple layers of security at Citibank: concurrent audit during application development, scenario based IT application controls and risk assessment by the concurrent auditors. Also, there is a need for real time detection of suspect transactions, even for smaller amount of transactions. Interestingly, the accused kept the withdrawals under $10,000 to avoid regulatory reporting and probably, they guessed it right that the bank would not look into transactions that would not require any regulatory reporting.
Most likely, the fraud detection predictive data models used by Citibank failed to notice such transactions or have assumed small amount transactions are relatively safer !!
Lessons learned ….
Banks need to strengthen their fraud detection data models.
All transactions, not just those required for regulatory reporting, need to be monitored.
Real time audit to be built into their security frameworks.
Needless to mention, build and continuously review appropriate application controls.
Information Security & Enterprise Risk Blog 