AI drives fraud prevention with behaviour-based anomaly detection

Fraud detection, data quality, and fostering trust are amongst AI’s top roles.

As fraud risks grow across industries, SAS Institute is advancing its use of artificial intelligence to enhance anomaly detection and mitigate fraud. Amir Sohrabi, Regional Vice President for ASEAN-Korea and Head of Digital Transformation for Emerging EMEA & AP, at SAS, shared insights into how AI innovation is addressing key challenges in fraud prevention and data analysis.  

Despite AI’s potential, developers face significant challenges, particularly in ensuring data quality. “The level and quality of data that organisations have access to have an impact on the quality of the models that are created,” Sohrabi explained.

Scaling AI models to accommodate growing datasets is another hurdle. “As you build out your models, are you able to make sure that they scale and perform as you infuse more data into those models?” Sohrabi asked.

To overcome these challenges and enhance productivity, Sohrabi stressed the importance of high-quality data access and trustworthy platforms. “We need to make sure there’s access to the right data. The quality of data is there, and then there needs to be a platform in which trust and performance are created,” he said.  

Sohrabi emphasised that developers need to ensure their models are transparent, accountable, and aligned with organisational goals, fostering trust within their teams and with external stakeholders.