Approximately 75% of Canadian banks are leveraging artificial intelligence (AI) for fraud detection purposes, according to a recent study by GFT. Additionally, around two-thirds of these institutions use AI to strengthen their cybersecurity measures, while nearly 68% apply it to improve customer service.

A Key Finding

Despite the widespread implementation of AI in various aspects, most banks report that their investments have yielded positive results. However, the greatest returns come from cybersecurity initiatives, with 23% identifying this as the area delivering the highest return on investment (ROI). Fraud detection follows closely at 22%, and automated customer support ranks third at 19%.

A Back-Office Focus

Another significant finding is that AI has played a larger role in back-office functions than in customer-facing initiatives. While most retail banks have invested in AI to enhance customer service, only 18% of them have observed measurable results in this area. In contrast, although fewer banks (one-third) have implemented AI for internal operational functions, the majority report that these applications deliver the greatest value.

The Double-Edged Nature

AI presents a double-edged sword for Canadian banks. Half of the surveyed institutions cite cybersecurity risk as their top challenge when considering AI adoption. As AI systems handle sensitive data and influence critical decisions, banks are increasingly balancing the need for innovation with the requirement for robust protection.

Seeking External Support

Currently, 63% of Canadian banks use AI in some form. They allocate more than one-third of their IT budget to this technology and plan to increase that investment by 20% over the next five years. Banks recognize that they will need support to fully harness AI’s potential. More than half are adopting a hybrid approach, combining in-house and outsourced teams, while an equal number prefer outsourcing their AI efforts to external partners as those building strong internal capabilities.

AI has become essential for fraud prevention and detection,” stated Suzanne Sando, Lead Analyst of Fraud Management at Javelin Strategy & Research. As attacks become more sophisticated and complex, AI and machine learning can detect anomalies and suspicious behaviors in a way that traditional technology cannot. The key is that these models improve over time, making fraud detection more precise as they mature.”