The Global Financial Industry Needs to Speed up Its Fraud Detection Through Analytics

Guy Muller, VP Engineering, RSD

Fraud and data theft is, unsurprisingly a significant headache for organizations, and particularly so for banks and insurance companies. Since 2005, there have been over 10 billion data records compromised in 8,000 breaches in the USA alone, that’s the equivalent of 1-2 a day.

While a major positive for many consumers and organizations around the world, the Internet has also created opportunities for fraudsters and data thieves, who work hard to steal from financial institutions. The answer seemingly would be for these organizations to simply put in place measures to spot and stop the fraudsters as they’re attacking. However, that is not always possible

In 2017 alone, intrusion attempts increased by 45% versus the previous year, and between 2012 and 2016 such attacks rose by 240%. The astronomical growth in these attacks demonstrates not only the need to identify them but to shut them down and stop future cases occurring.

However, the problem doesn’t just lie in the theft of the data itself, but in what the criminals do with it, namely cloning identities. With access to such large amounts of stolen personal information, fraudsters are able to adopt personas and attempt to carry out illegal transactions, including taking control of bank accounts that aren’t theirs.

But, in today’s hyper-connected world, consumers demand instant access to their accounts and insurance policies, and those financial institutions that don’t offer it will lose out on valuable revenue, despite the multitude of channels it also creates that thieves can exploit. As a result of what is effectively a series of open doors, it’s difficult to accurately assess the costs associated with these fraudulent activities, though they are known to be extensive.

It’s therefore imperative that the institutions make use of their full arsenal to fight the fraudsters to protect themselves and their customers. One way of doing this is by using analytics and business intelligence solutions that can help teams to identify activity that doesn’t look right. In organizations that are 100% cloud-based, accessing the data is easier and helps stop the criminals in their tracks. However, given that 70% of Fortune 500 companies, 92 of 100 of the largest banks and all top 10 of the largest insurance organizations rely on legacy mainframe systems, access to data on these systems ideally needs to be just as fast so they can respond to threats quickly. But that’s generally not possible. The reason for this is that, while incredibly powerful and highly secure platforms, mainframes are purposefully built as silos that are not connected with the rest of the IT infrastructure, and mainframe specialists and data scientists don’t speak the same language.

With the data that’s required to identify these threats stored on the mainframe, the logical option would be for the BI services to be as well, but that is often counterproductive. By extracting the data from the mainframe’s closed architecture, and performing analytics from outside alongside other data sources, BI teams can build more comprehensive and accurate models that give them a true picture of any potential fraudulent activities.

A solutionis to use technical offeringsthat enablethe transfer ofmission critical data without impactingeither the security or the quality of service of themainframe applications. The latest solutions capture data from any mainframe application and use the SMF logging mechanism to transfer the data to an open system (for example Linux). This data is then made available as a JSON file, a format that can be utilized by any BI solutions.

This three-stage mechanism ofcapture, transfer, andextraction has many advantages, the most obvious being that, by delivering the data into a service that analyst teams already use, anyone in the fraud team can work with the files to spot and stop the fraudsters. An additional bonus is that, once set up, its maintenance requires only very limitedmainframe skills, a definite advantage because the pool of seasoned professionals of this architecture is getting smaller and smaller as experts retire without new entrants to match. And finally, this type of solution does not generate a mainframe overhead by only capturing and transferring the required data to the BI platforms.

By having such access available quickly and efficiently, financial institutions have an opportunity to potentially tap into the mainframe at will, meaning they’re much more likely to halt fraudsters in their tracks. With fraud and data access becoming a higher priority amongst consumers, particularly in the context of GDPR coming into effect in Europe, every organization around the world needs to take their data security seriously, and that has to include having fast and secure access to their critical information. Those that fail to do so will not only fall foul of the criminals, but the law courts and their customers too.

Guy Muller, VP Engineering, RSD

Fraud and data theft is, unsurprisingly a significant headache for organizations, and particularly so for banks and insurance companies. Since 2005, there have been over 10 billion data records compromised in 8,000 breaches in the USA alone, that’s the equivalent of 1-2 a day.

While a major positive for many consumers and organizations around the world, the Internet has also created opportunities for fraudsters and data thieves, who work hard to steal from financial institutions. The answer seemingly would be for these organizations to simply put in place measures to spot and stop the fraudsters as they’re attacking. However, that is not always possible

In 2017 alone, intrusion attempts increased by 45% versus the previous year, and between 2012 and 2016 such attacks rose by 240%. The astronomical growth in these attacks demonstrates not only the need to identify them but to shut them down and stop future cases occurring.

However, the problem doesn’t just lie in the theft of the data itself, but in what the criminals do with it, namely cloning identities. With access to such large amounts of stolen personal information, fraudsters are able to adopt personas and attempt to carry out illegal transactions, including taking control of bank accounts that aren’t theirs.

But, in today’s hyper-connected world, consumers demand instant access to their accounts and insurance policies, and those financial institutions that don’t offer it will lose out on valuable revenue, despite the multitude of channels it also creates that thieves can exploit. As a result of what is effectively a series of open doors, it’s difficult to accurately assess the costs associated with these fraudulent activities, though they are known to be extensive.

It’s therefore imperative that the institutions make use of their full arsenal to fight the fraudsters to protect themselves and their customers. One way of doing this is by using analytics and business intelligence solutions that can help teams to identify activity that doesn’t look right. In organizations that are 100% cloud-based, accessing the data is easier and helps stop the criminals in their tracks. However, given that 70% of Fortune 500 companies, 92 of 100 of the largest banks and all top 10 of the largest insurance organizations rely on legacy mainframe systems, access to data on these systems ideally needs to be just as fast so they can respond to threats quickly. But that’s generally not possible. The reason for this is that, while incredibly powerful and highly secure platforms, mainframes are purposefully built as silos that are not connected with the rest of the IT infrastructure, and mainframe specialists and data scientists don’t speak the same language.

With the data that’s required to identify these threats stored on the mainframe, the logical option would be for the BI services to be as well, but that is often counterproductive. By extracting the data from the mainframe’s closed architecture, and performing analytics from outside alongside other data sources, BI teams can build more comprehensive and accurate models that give them a true picture of any potential fraudulent activities.

A solutionis to use technical offeringsthat enablethe transfer ofmission critical data without impactingeither the security or the quality of service of themainframe applications. The latest solutions capture data from any mainframe application and use the SMF logging mechanism to transfer the data to an open system (for example Linux). This data is then made available as a JSON file, a format that can be utilized by any BI solutions.

This three-stage mechanism ofcapture, transfer, andextraction has many advantages, the most obvious being that, by delivering the data into a service that analyst teams already use, anyone in the fraud team can work with the files to spot and stop the fraudsters. An additional bonus is that, once set up, its maintenance requires only very limitedmainframe skills, a definite advantage because the pool of seasoned professionals of this architecture is getting smaller and smaller as experts retire without new entrants to match. And finally, this type of solution does not generate a mainframe overhead by only capturing and transferring the required data to the BI platforms.

By having such access available quickly and efficiently, financial institutions have an opportunity to potentially tap into the mainframe at will, meaning they’re much more likely to halt fraudsters in their tracks. With fraud and data access becoming a higher priority amongst consumers, particularly in the context of GDPR coming into effect in Europe, every organization around the world needs to take their data security seriously, and that has to include having fast and secure access to their critical information. Those that fail to do so will not only fall foul of the criminals, but the law courts and their customers too.