SAS study: 8 out of 10 experts want to combat financial fraud with the help of generative AI by 2025

A global survey of anti-fraud professionals by SAS, a global leader in AI and analytics, in partnership with ACFE, reveals a high level of enthusiasm for generative AI, but also presents a challenging reality in the field. Generative AI has captured the public’s imagination with its capabilities and promises to integrate into every aspect of society. It makes sense, then, that 83% of anti-fraud professionals anticipate adding this technology to their arsenal in the next two years.

This is the main conclusion of the anti-fraud technology study conducted by SAS, together with the Association of Certified Fraud Examiners (ACFE). The 2024 Anti-Fraud Technology Benchmarking Report is the third global study by SAS and ACFE since 2019. Its latest edition brings to the forefront insights from approximately 1,200 ACFE members, interviewed at the end of 2023. The data obtained show clear trends in the evolution of the anti-fraud fight, starting from 2019: Among them, it stands out:

Interest in artificial intelligence (AI) and machine learning (ML) is higher than ever.

Currently, about 1 in 5 anti-fraud professionals (18%) cite the two tools among those they use to detect fraud. Another 32% anticipate implementing such technologies within the next two years. At this rate, the use of AI and ML in anti-fraud programs will triple by the end of 2025. However, the adoption rate of artificial intelligence and machine learning is below expectations. Although interest is very high, the actual adoption of AI and ML has increased by only 5% since 2019. Estimates in 2019 and 2022 predicted rates of 25% and 26%, respectively.

While the use of a large number of data analysis techniques has reached a plateau, the use of biometrics and robotics in anti-fraud programs has been in continuous growth and development. The use of physical biometrics has increased by 14% since 2019, and is now a resource for 40% of respondents.

One in five (20%) respondents use robotics, including robotic process automation, up from 9% in 2019. These technologies are most common in banking and financial services. 51% of respondents use physical biometrics and 33% use robotics.

“The accessibility of generative AI tools makes them especially dangerous if they fall into the wrong hands,” says John Gill, president of ACFE. “Three out of five organizations expect an increase in budgets for anti-fraud technology in the next two years. How they invest these funds will determine who will have the upper hand in what has become a technological arms race in which the adversaries are the criminals. It’s an uphill battle because unlike fraudsters, organizations need to use new technologies ethically.”

“The huge interest in advanced analytics techniques, juxtaposed with their modest adoption rate proves the complexity of scaling AI and analytics solutions,” says Stu Bradley, Senior Vice President of Risk, Fraud and Compliance Solutions at SAS. “At the same time, this underlines the importance of choosing the right technology partner. AI and machine learning are not simple, plug-and-play applications. However, their benefits can be easily achieved by delivering modular solutions across the spectrum of risk management, based on a unified platform based on artificial intelligence. This is the SAS approach, through the SAS Viya platform.”

The Future of Generative AI: Success or Failure?

Will the integration of generative AI into anti-fraud programs reach the level predicted by the increased interest of survey respondents? Or will real challenges such as financial limits, data quality and skills shortages hinder future growth? Only time can answer these questions – but companies must move towards such technological solutions. Responsible innovation requires organizations to ask themselves not just “can we do this?” but also “should we do this?”

“The use of generative AI in anti-fraud initiatives could have a significant role in identifying anomalies, trends or indications from large volumes of data, with minimal use of resources,” said one respondent of the study. “However, the organization will need to ensure that clear and fair rules are in place to prevent errors and bias.” “Generative AI has made major strides in recent years, so it’s no wonder organizations are incorporating it into anti-fraud initiatives,” said Mason Wilder, Research Director at ACFE. “At the societal level, we are still discovering what are the advantages and disadvantages of using this technology, but more and more organizations are taking that first step. It will be interesting to see how quickly this new technology will be adopted, but also how much it will develop in the future.”