Home | Tech | As fraud increases and becomes harder to detect, 20% don't use analytics to fight it

As fraud increases and becomes harder to detect, 20% don't use analytics to fight it

By
Font size: Decrease font Enlarge font
As fraud increases and becomes harder to detect, 20% don't use analytics to fight it

It’s not just an IT issue anymore.

Fraud is expected to increase during the next year (24.9%) and became tougher to detect during the past year (31.4%), according to a recent survey by professional services firm Deloitte. Yet, one-fifth of respondents (20.8%) don't use analytics or rare event modeling in their anti-fraud programs. 

"Fraudsters are getting smarter, but tools to identify schemes earlier are advancing, too."

"Fraudsters are getting smarter, but tools to identify schemes earlier are advancing, too," said Shuba Balasubramanian, Deloitte Advisory principal in advanced analytics, Deloitte Transaction and Business Analytics LLP.  "We see increasingly more organizations using advanced analytics for rare event  modeling.  Identifying anomalies — possibly fraud, waste or abuse — earlier than ever before can help stem financial losses and negative brand impact."

Kirk Petrie, Deloitte Advisory principal in advanced analytics, Deloitte Transaction and Business Analytics LLP, added, "When done well, analytics and rare event modeling accelerate fraud prevention and detection efforts. Good execution often involves executive leadership, legal, finance, risk, compliance — along with data scientists — so that faint signals of fraud can be identified and organizational improvements made.  Unfortunately, few companies currently involve data science in their compliance directorates."

 

Findings indicate the biggest challenges to using fraud analytics effectively are lack of resources (29.6%), treatment of fraud analytics as an IT issue, not an organizational priority (15.1%) and inability for leadership to understand the value (10.5%).

Join PRESIDENT&CEO on LinkedIn

Subscribe to comments feed Comments (0 posted)

total: | displaying:

Post your comment

  • Bold
  • Italic
  • Underline
  • Quote

Please enter the code you see in the image:

Captcha