How AI Can Be Used to Perpetuate a Fraud—and How Fraud Examiners Pull Back the Curtain
First: A History Lesson
In January 1920, an Italian immigrant in Boston started a company that engaged in arbitrage involving a form of postage called “international reply coupons.” The man convinced acquaintances in his community to invest money in his company with a promise that he could double their money within 90 days. The plan was physically infeasible, requiring millions of coupons when less than 30 thousand were in circulation. The man behind this fraudulent scheme, Charles Ponzi, would lie to investors about the returns. He repaid old investors with money obtained from newer investors. This scheme quickly fell apart, and he would plead guilty to mail fraud by November of 1920, less than a year after he began.
For a more recent example, let’s look at convicted fraudster Bernie Madoff. Madoff, like Ponzi, used his personal reputation to obtain investments from the wealthy communities that he was a part of. Madoff couldn’t achieve the returns he promised, and he needed to entice investors to keep their money with him rather than cashing out. To perpetuate the lie, he had a team of employees who were directed to use computers to falsify account statements and trading records using backdated information.
As technology advances, fraudsters continue to find new ways to enhance their operations. Artificial intelligence (AI), when misused, can make fraud schemes more convincing, scalable, and harder to detect. Sophisticated AI can aid in generating the illusion that Bernie Madoff and his team worked hard to perpetuate. However, AI also provides powerful tools for fraud examiners. Here’s how AI both facilitates and helps uncover these schemes.
How AI Enhances Fraud Schemes
Fraudsters can use AI to help scale up the scheme. No longer does it need to be limited to an in-group that knows the fraudster personally. AI-generated videos or audio can be used to impersonate executives, adding credibility to their scheme without necessitating a personal relationship with the in-group. AI-powered tools can analyze social media and other data to identify high-value targets and tailor pitches to their interests or vulnerabilities. And instead of a team of humans reaching out to targets, AI chatbots can engage with potential investors, answering questions in real-time to build trust and address concerns.
Additionally, rather than hiring someone to falsify records, AI can automate the creation of fake financial reports and performance data to convince investors of profitability. As AI becomes more sophisticated, it could be used to generate realistic “real time” dashboards mimicking legitimate investment platforms showing false returns obscured by complex transactions. Like with Madoff, the goal is that victims continue to reinvest their funds rather than cashing out.
Although AI can help add credibility to claims, the method of avoiding a fraudulent scheme, at its core, has not changed. Investors should always be exceedingly wary of “too good to be true” returns. Don’t trust the random message you get on WhatsApp or Discord touting the huge opportunity to invest in Cryptocurrency. Don’t be the “greater fool” left holding the bag while the fraudsters cash out of their pump-and-dump.
New technologies like AI simply make it easier for anyone to run a fraud, and it becomes a numbers game. If the fraudsters reach out to enough people, eventually someone will make a mistake and take the bait. What happens then?
How Fraud Examiners Use AI to Uncover Fraud Schemes
Luckily, fraud examiners can also use AI tools to help investigate and expose financial fraud. Fraud examiners work with teams of law enforcement and lawyers to put the fraudsters in jail and recover as much of their ill-gotten gains as quickly and as easily as possible.
Fraud examiners still rely on source documents like bank statements, invoices, or other publicly available market data. For many years, fraud examiners have been using advanced data analytics tools to process and export data from source documents to aid in analysis. Now, AI analysis is built into those same tools, giving rise to new and more efficient ways of analyzing financial patterns.
AI-powered forensic accounting tools analyze vast amounts of transaction data to identify inconsistencies, such as returns that are too consistent or cash flows that don’t match claimed investments. Algorithms track the movement of funds through multiple layers of obfuscation, and AI can also help prepare visualizations of those movements to report our findings to interested parties.
AI can also help identify red flags when it comes to falsified documents. Fraud examiners use AI tools to cross-check reported returns against actual market data, exposing fabricated performance metrics. AI can also identify patterns in fraudulent accounts, such as repeated use of similar data points across multiple identities.
Staying Ahead of AI-Powered Fraud
Let’s not overstate AI’s abilities. In its current state, Artificial Intelligence is more “illusion” than “intelligence.” Like in the Wizard of Oz, there’s a human charlatan pulling its levers. With the AI tools mentioned above and a little bit of human incredulity, we can pull back the curtain and expose the fraudster for who they truly are.
At Ahuja & Consultants, we are on the cutting edge of AI-driven forensic accounting and fraud detection techniques to uncover even the most sophisticated schemes. If you suspect fraudulent activity or need assistance evaluating an investment, contact us today. Together, we can protect your financial future and bring transparency to an increasingly complex world.
