The Interplay of AI and Fraud in the Economy: The Good, the Bad, and the Unethical
Updated: Jul 25
An Uncertain Economy
As we enter the third quarter of 2023, many economic forecasts are highlighting the instability of the global economy. Continued geopolitical tensions, elevated long term core inflation, and the recent regional banking turbulence in the U.S. have concerned many financial analysts, and they hint of danger to come. The International Monetary Fund warns that economic growth remains low by historical standards, and financial risks continue to rise. While the future is not certain, it is of utmost importance that financial institutions be aware that an unstable or slowing economy can contribute to an increase in fraud.
Heightened Risk of Fraud
During the financial crisis of 2007-2010, the Association of Certified Fraud Examiners conducted a survey of over 500 randomly selected Certified Fraud Examiners (CFEs) who specialized in fraud detection, prevention, and deterrence. The results of the survey strongly suggested that the intense financial pressures of the 2007-2010 economic crisis led to an increase in fraud. In addition, nearly half (49.1%) of the respondents reported financial pressure as the biggest contributing factor to an increase in fraud, in consideration with increased opportunity (27.1%) and increased rationalization (23.7%). The fraud triangle, set forth by Donald Cressy, describes these three factors that are present in every fraud - pressure (need for money, financial struggles, etc.), rationalization (the mindset of the fraudster that justifies them to commit fraud), and opportunity (the situation that enables fraud to occur).
Pressure - Pressure to meet goals, deadlines, demand for increased productivity, fear of losing job, or reduction in salary could pressure an employee to act out against their company.
Rationalization - An employee might justify their actions by thinking that the company can afford the losses set forth by the fraud, that their struggles merit some additional support, their goals are unattainable, or as a revenge scheme.
Opportunity - Times of economic hardship can create many more opportunities for fraudsters to act on. Employee cuts, lack of adequate supervision, or the promotion of under-experienced people can open up the window for fraud.
In a period of economic challenge, these three factors are often at significantly higher levels. As a result, challenging times can lead to higher levels of fraud. Currently, fraud levels have been steadily rising since 2020. Digital fraud losses are expected to surpass $343 billion globally from 2023 to 2027. Socioeconomic challenges and an increase in technology have encouraged and enabled more fraudsters to engage in criminal behavior.
Advantages of AI in Fraud Reduction
One modern development that can be used to combat the increase in fraud is the recent improvements to Artificial Intelligence (AI). Progressive technological advances have allowed lenders to gain a competitive advantage over fraudsters and stop crime before it happens. One way this can be accomplished is through the optimization of identity verification systems. AI can be used to analyze multiple points of data and biometric information at once, providing a multi-layered and efficient approach to the verification of data. Under this multi-layered approach, the authentication system can provide an effective identity verification service even if one method of authentication is compromised. Furthermore, the analytic capabilities of an AI program are not limited to identity verification. The possibilities are virtually endless and can be used to prevent fraud in numerous ways, including the verification of important documents, and improvements to Know Your Customer (KYC) and Anti-Money Laundering (AML) technology. At the moment, 77% of banks are leveraging AI for fraud protection, with a 90% average improvement in fraud detection rates achieved by AI solutions. With an accurate method of data analysis, banks/financiers can be confident that they are entering a safe deal.
Banks/Financiers Won’t Want to Miss
SEON - “SEON analysis tools can extract data from a single email address, phone number, or ip address. All that data is fed through customizable risk rules, which can also be suggested via a whitebox machine learning algorithm. SEON gives banking risk managers an extra layer of data to gain full confidence in accepting or rejecting risky customers.”
SAS - “Their framework is designed for anomaly detection to reduce risk and to consolidate your views of potential fraud. It is widely used in the BFSI sector (banking, financial services, and insurance), as it enables enterprise-wide monitoring from a single platform.”
LexisNexis Risk Solutions - “In terms of risk rules, LexisNexis Risk Solutions offers anything you may need, from behavior tracking to custom rules and machine learning suggestions – all available via API calls or on-premise installation.”
Feedzai - “Trusted by banks such as Citi, Lloyds, and Santander, Feedzai is dedicated to protecting financial institutions in three ways: securing account openings, controlling transaction fraud, and stopping money laundering.”
Experian Hunter - “As Experian rightly puts it, the first and best strategy to reduce fraud losses is to identify criminals at the point of application. Their detection method is designed to be highly configurable, whether you’re a traditional retail bank or a challenger bank with low-friction needs.”
Comply Advantage - “ComplyAdvantage includes transaction monitoring, customer screening using sanction lists and PEP lists, and even checks based on the Financial Action Task Force to monitor adverse media.”
iComply - “iComply, also known as iComplyKYC, focuses on due diligence at the banking onboarding stage. It’s an all-in-one KYC and AML solution that can also help you screen for identity fraudsters at signup. When it comes to risk management, you will get all the standard KYC and AML features such as watchlists, PEP, and adverse media checks, as well as automation tools to save you time when processing customer data.”
Ondato - “Ondato is a full KYC and AML solution that is designed to help with onboarding, business intelligence, and user base management.”
ACTICO - "Actico is the trusted solution by companies such as Santander, Volkswagen Financial Services and ING, among others. You get KYC tools using risk scores and AML features such as PEP and sanction lists screening. If you’re a traditional bank still struggling to compete with neobanks, Actico also offers digital transformation services too.”
Ocrolus - “Detect altered identity documents, invalid dates, abnormal fonts, irregular formats, and more with features powered by intelligent AI technology.”
Bankers Beware: AI-Assisted Fraud
While AI can be a formidable resource in the reduction and prevention of fraud, technological developments can also lower the barrier of entry to fraud and create new ways to carry out fraud schemes against banks. In the past, financial fraud was a challenging endeavor, and required a certain amount of knowledge and experience to carry out. With the development of recent tools and programs, fraudsters have it easier than ever. AI has the ability to create false or misleading information on a large scale, automate scams, and aid in impersonation and/or social engineering. For example, Financial IT has stated that new AI software has already allowed scammers to produce fake invoices, documents, and financial statements that are often more personalized and accurate than the sources they are impersonating. This could have an enormous impact on the equipment finance industry, in which two common types of fraud include transactional and asset-leasing fraud. In addition, AI tools which will allow fraudsters to create realistic images, video, audio, and 3D models will soon be broadly available to the general public. These tools could enable the production of hyper-realistic “deepfake” documents such as passports, identification cards, and other official papers which could even bear official stamps and traditional hallmarks used to identify authentic forms of identification. This would make it much more difficult to discern between what is real and what is fake.
Factors that encourage the use of AI in fraud include…
Speed and efficiency - AI’s ability to process large amounts of data in a short amount of time makes it a useful tool for the automation of fraudulent activities.
Anonymity - AI can be used to carry out fraudulent activities without it being traceable to the fraudster.
Increased sophistication of attacks - AI can be used to specifically adapt to the defenses of targeted organizations.
Finance companies should ensure that they remain informed on the latest sophistications of AI-based fraud trends. By maintaining strong policies and security procedures, they can stay ahead of the trend and stop fraud before it happens.
AI can also pose potential ethical and regulatory concerns. The biased results that are present in popular search engines, many of which AI relies upon, can reflect upon the responses that AI gives. Consequently, data may have embedded racial or gender bias. Furthermore, the process through which AI makes its decisions is not always clear to the outside observer, which can contribute to a lack of transparency in decision-making processes. There are also concerns about methods of data collection, user privacy, and the overall security of the system.
It is imperative for banks and lenders to consider how an issue with ethics could harm the company. Unethical behavior is visible and important to customers. If word of unethical practices reaches the public, the company risks high levels of scrutiny and a potential decline in customer loyalty. This could also lead to a loss of sales and fall in share prices. In an increasingly competitive business environment, banks/lenders must act ethically to avoid permanent damage to their company.
Although AI can be used as a powerful tool for fraud prevention, finance companies must also consider the ways it can be used against them to commit fraud. Moreover, there are various ethical concerns that can come with the use of AI. Banks, lenders, and finance professionals should take these factors into account when implementing fraud-related policies and procedures.
The author, Tenor D. Ickes, was born in Wurzburg, Germany, to U.S. military parents serving abroad. He is a summer intern at Oswald Law Firm, where he enjoys writing on trending business, banking & finance law, and litigation topics. He earned a B.A. from San Jose State University while working as a law clerk at a Silicon Valley banking litigation firm. He is an incoming 2L student at UC College of the Law San Francisco where he is pursuing a Juris Doctor degree, and he is a student member of the San Francisco Bank Attorneys Bar Association and The Bar Association of San Francisco. In his free time, Tenor enjoys spending time with family, international travel, and writing music.
The above article provides information only and does not create an attorney client relationship. It should not and cannot be construed as legal advice. Need help with fraud litigation? Oswald Law Firm can help. Contact Harmony Oswald, Esq. at email@example.com
Copyright 2023 Oswald Law Firm. All Rights Reserved.