AI technology can lend a hand in corporate fraud investigations

Rising rates of corporate fraud are causing businesses in the GCC to carry out an increased number of costly investigations into misconduct, with 97 per cent of compliance experts believe fraud has a ‘significant impact’ on their organisation.

Although the GCC has achieved sustained economic growth for years, the corporate investigations landscape has struggled to keep pace with the risk Gulf-based companies are facing from fraudsters. Companies will need to adopt more advanced technology to keep control of the costs of that rising burden of fraud investigations. Without the aid of technology, businesses may find the costs so prohibitive they are disincentivised from undertaking thorough investigatory work.

History suggests that periods of economic disruption typically result in higher instances of fraud and consequently an increase in fraud investigations. There was a dramatic increase in fraud investigations following the 2008 financial crisis, and indications are the same this is happening following the pandemic.

The introduction of lockdowns led to a huge number of changes in the way businesses operated, which opened the door to increased fraud. Managers were unable to monitor employees working from home as effectively as in the office, while in some cases the sudden switch to digital working opened holes in companies’ IT infrastructure that fraudsters and cyber criminals were able to exploit.

Threat of fraud

As with any time of increased economic stress, temptation for some to commit fraud increased. Targets became harder to reach, potentially putting bonuses at risk. The threat which the pandemic posed for the viability of businesses may have tempted some to bend the rules.

Increasing levels of fraud by employees since the pandemic has been a global problem. While great strides have been made in addressing fraud risks in the region, especially regarding data protection and AML issues, many businesses still lack the technological infrastructure to properly investigate instances of fraud. This makes them more vulnerable to financial misconduct by employees.

Certain businesses are more vulnerable to specific types of fraud. For instance, the energy and construction sectors need to be especially vigilant to the risk of procurement fraud. This is where a tender process has not been properly followed, potentially resulting in rigged bidding, or where goods or services delivered were inferior to those specified in the order.

Bring investigation costs down

Dealing with fraud investigations can cost businesses significant sums and take up a huge amount of valuable employee time. Even a small internal fraud investigation can easily lead to a cost of around $100,000, if traditional, manual methods are employed.

For example, our Dubai office is now leveraging Reveal-Brainspace, provider of an AI-powered e-discovery and investigations platform, to maximise the speed and cost-effectiveness of its investigation services. Reveal uses AI to dramatically speed up investigations by quickly processing emails and other documents to identify red flags and bringing together all the evidence in an investigation far more quickly than human investigators could.

A traditional, manual review of 300,000 documents would require a total of 4,800 hours to be spent on the investigation, involving up to 60 staff. This can be cut to just 1,200 hours and 15 staff and achieve more accurate results.

Getting AI to handle it

AI-driven investigations are not only much more cost-effective and less time-consuming, they are also much more thorough, making links between disparate documents and building up the picture of a fraud case far more effectively than human investigators acting without technology. Experienced investigators, who are well-versed in the facts of the investigation, are able to train and configure AI tools to quickly map the communication network of targeted individuals.

This allows investigators to quickly build a picture of their communications and identify ‘individuals of interest'. This means that it is much harder for the perpetrators to cover their tracks, and much less likely that the investigation will miss the ‘smoking gun’ which proves fraud has taken place.

Among other things, AI can highlight concerning language in emails which may suggest an individual is being pressured to act improperly. Communications at usual times such as the weekend can be easily logged. Significant numbers of deleted emails or the sharing of privileged information to an external domain can be instantly flagged.

GCC-based businesses, like their counterparts worldwide, must ensure these investigations are as cost-effective as possible, with time and resource used internally kept to a minimum.