Artifical intelligence: robot hand and human hand

Leading organizations are increasingly integrating human and machine intelligence for forensic investigations.  By combining technological analysis—including artificial intelligence (AI), machine learning and statistical concepts of cognitive analytics—with deep investigator understanding of fraudster motives and methods, many are identifying, investigating and thwarting fraud schemes more effectively and efficiently than ever before.  But, how mature is your organization’s approach to forensic investigations?  Deloitte’s Satish Lalchand discusses in the first of a series of articles on the future of forensics.

As research from the Association of Certified Fraud Examiners (ACFE) has shown consistently over the years, organizations lacking antifraud controls suffer twice the median fraud losses—typically five percent of annual revenue—of those with controls in place.

While technology challenges abound—ranging from over reliance on rules-based testing to proliferation of unstructured data to organizational “information silos,”—leading organizations are increasingly integrating human and machine intelligence for forensic investigations.

By combining technological analysis—including artificial intelligence (AI), machine learning and statistical concepts of cognitive analytics—with deep investigator understanding of fraudster motives and methods, many are identifying, investigating and thwarting fraud schemes more effectively and efficiently than ever before. This integration of human and machine intelligence is the hallmark of a mature forensic investigation program.

For compliance, finance, forensic, risk, legal and other executives, here are some questions that can help discern how mature their organizations’ forensic investigation programs are:

  1. How integrated is your organization’s data? Integrating structured and unstructured data from across the enterprise and data from external sources like watch lists and social media can paint a much broader picture of activities and transactions, which experienced forensic investigators can piece together with fewer false positives. Many organizations struggle with ongoing data integration, as data volumes and types continue to proliferate. But, the most mature organizations endeavor to combine and analyze available data for multiple uses, including forensic investigations.
  2. How holistic is your organization’s approach to fraud analytics? How frequently are analyses run to detect and prevent fraud and corruption? Is the data analyzed within business units or other silos, or in an enterprise-wide manner? If functions like customer experience management and supply chain have strong analytics operations in your organization, could those teams help contribute analyses to forensic investigations as needed? The most mature forensic investigation programs take a holistic approach, constantly honing efforts to detect and prevent fraud.
  3. Does your organization risk score transactions or entities? Transactions don’t commit fraud, people do. It sounds simple enough, but shifting away from transaction-focused risk ranking to entity and behavioral pattern risk ranking can be tough for some organizations. Data-driven advanced analytics models incorporating text analytics and network analysis can enable organizations to rank risks at the individual or entity level, rather than the transaction level. Organizations that focus their risk-ranking efforts on entities and the patterns of human behavior tied to them can help improve ranking accuracy and efficiency, another aspect of a more mature forensic investigation program.
  4. Are predictive tools leveraged in your organization’s fraud analytics program? Just as integrated unstructured and structured data, holistic analytics and entity-based risk ratings are hallmarks of mature forensic investigative programs, the use of predictive analytics is also gaining ground. Machine learning, cognitive computing and other advanced analytics techniques are designed to enable investigators to identify consistent traits of bad actors that can then be scanned to identify similar “digital footprints” of such behavior in real time, predicting when the next scheme is underway. Few organizations have reached this stage of forensic analytics maturity today, but we expect it to be commonplace in coming years.

By employing advanced analytics approaches in combination with field-demonstrated forensic techniques, organizations can better detect, isolate, and deter fraud attacks, with potentially significant positive impact on an organization’s performance and productivity.


Satish Lalchand

Satish Lalchand is a Deloitte Risk and Financial Advisory principal in forensic analytics, Deloitte Transactions and Business Analytics LLP.  A certified fraud examiner (CFE), he specializes in anomaly detection and data analytics, business rules development and predictive modeling. Lalchand has in-depth knowledge of fraud rules and model creation for prevention, detection and investigation with a broad range of experience in managing and leading engagements in these areas. He can be reached at [email protected].

Related Post

Got Compliance News?

We do!  Sign up for CCI’s free weekly eBlast to get GRC news, views, jobs & events delivered to your inbox once a week.  Cancel anytime.

Click to Subscribe.