How Forensic investigators gain an edge with AI

Switzerland is renowned for its business-friendly environment and immense political stability, which has made it a successful international centre of business. Yet this also attracts a sophisticated set of corporate criminal activities. Using the latest AI technology in investigations however can enhance the ability to identify and investigate attacks, and help investigators get to the root cause quicker. Moreover, it can improve detection and prevent re-occurrence.

Why gaining the edge is important

The information age has led to unbounded criminal ingenuity. Data now needs to be fiercely protected alongside intellectual property. Companies, especially those in Switzerland, must adapt rapidly by installing more sophisticated controls and monitoring technologies. If they lack the best anti-fraud controls, they are worse off, suffering twice the median in fraud losses, compared to those with controls in place.

Combining people and AI in a forensic investigation can give a company the edge:

• It introduces automation, which saves significant time and expense, and allows investigators to focus more on where fraud might occur.
• It helps companies detect criminal activity from the vast amounts of unstructured data they have collected, such as from videos, images, emails, and text files.
• It is a more dynamic approach than rule-based testing, which is limited to monitoring fraud risk across a single data-set.
• It gets rid of the information silos that can build up, which can further impede an analytics-aided investigation: this occurs when locally-tailored processes prevent integrated data sharing, which creates barriers to an investigation.

How the forensic investigators do it

There is a temptation during an investigation to rely on previous experience and knowledge, through an intuition-driven approach. An experienced forensic investigator must look ahead and not behind for guidance. The amount of data that must be analysed is not only increasing, but its nature and how you interpret it, is constantly changing. This only serves to amplify human biases.

A forensic team therefore needs to run an integrated analytics-driven investigation.

Here is how it works:

• They first look at how capable a company is at detecting fraud and carrying out forensics by determining where that company lies on a maturity model that captures the people, the processes and the tools used to detect fraud.
• Then they integrate structured and unstructured data from internal and external sources into risk models that are important for carrying out advanced analytics.
• Data-driven advanced analytic models, which incorporate text analytics and network analysis, are then used to rank risks at a company-level, rather than at a transaction level.
• Advanced analytics techniques, such as machine learning, and cognitive-data analytics are then finally applied.

Cognitive-data analytics, which is self-learning, allows data to be digested dynamically and in real time. Data mining takes place, patterns are seen and recognised, and natural language is analysed. These are all processed together, much in the same way as the human brain operates. This is how forensic investigators can gain an edge during an investigation.