Financial institutions have long used AI to combat fraud. AI tools have reached a point where they can detect fraud in real time. These improvements have greatly decreased the number of potential frauds committed against customers and banks.

However, criminals are now using AI to beat AI fraud detection tools. CNBC reported, “since the fourth quarter of 2022, there’s been a 1,265% increase in malicious phishing emails, and a 967% rise in credential phishing, according to a new report by cybersecurity firm SlashNext.”

AI is being used to generate phishing emails (imposter emails that look like real emails) to gain access to customer accounts and bank data. Such AI tools allow fraudsters to commit crimes at scale. Phishing emails can be sent out in mass and written in the native language of the target country. Additionally, these emails can be written in the target bank’s tone, all using GenAI (Generative AI).

Larger financial institutions (FIs) have an advantage over smaller ones when training anti-fraud AI models. This is because of the large amount of data at bigger FIs compared to smaller FIs. Data sharing could help FIs overall. Unfortunately, this is not common practice in the industry.

To improve its fraud-fighting efforts, Visa has launched a new set of AI fraud-fighting tools to protect issuers, merchants, and consumers, even if transactions did not occur in the Visa network. To provide this expansion of its anti-fraud services, Visa is utilizing neural networks and deep learning models.

There are three main areas that Visa is providing updates for:

• Transaction risk scoring
• Advanced Authorization and Risk Manager for non-Visa card payments
• Real-time Account-to-Account Payment Protection

These updates will be available to Visa customers during the first half of this year. Visa said these solutions should also help smaller FIs that lack the resources or teams to manage anti-fraud AI models.