North American banks struggle with deepfakes, social engineering

Financial institutions in North America face a spike in fraud schemes, with social engineering scams increasing tenfold in the past year. According to BioCatch data, digital banking fraud has evolved, leveraging advanced tactics like deepfake media to deceive both customers and financial institutions.

BioCatch’s latest insights reveal a shift in fraud tactics. Social engineering scams, which manipulate victims into divulging sensitive information, now account for a notable share of digital banking fraud.

These schemes often exploit trust through impersonation and emotional manipulation, including fake customer service calls or text messages. The growing accessibility of generative AI tools has fueled the rise of deepfake technologies, enabling fraudsters to mimic voices or create realistic visuals to deceive their targets.

“As we outlined in our 2024 AI, Fraud, and Financial Crime Survey and ScamGPT white paper, artificial intelligence is super-charging fraud, compounding its impact, and allowing bad actors to scale and sophisticate their scams with deepfakes and other devices,” says BioCatch global advisory director Seth Ruden.

“As the industry deploys the newest authentication methods in both account opening and account takeover processes, fraudsters will undoubtedly attack these as well.”

The Financial Crimes Enforcement Network (FinCEN) issued an alert on fraud schemes exploiting deepfake media to compromise financial systems. These attacks are often part of larger, coordinated efforts that target high-value transactions or account takeovers.

Similar to BioCatch, FinCEN has also observed a rise in suspicious activity reports from financial institutions, highlighting the suspected use of deepfake media. These reports frequently point to the creation of fraudulent identity documents designed to bypass identity verification and authentication processes.

Financial impact and industry response

The scale of these attacks is reflected in Sift’s Fraud Industry Benchmarking Resource (FIBR), which highlights metrics such as transaction fraud rates and the cost of account takeovers.

Within the Sift Global Data Network, credit and debit cards are implicated in 85 percent of fraudulent transactions. However, the trend shifts in specific industries, particularly those characterized by high transaction volumes.

In the iGaming and online gambling sector, credit and debit card fraud accounts for a lower share at 64 percent. Instead, electronic fund transfers make up 20.5 percent of fraudulent activity, while digital wallet payments contribute 15.5 percent.

The report warns that businesses failing to adapt to evolving fraud risk substantial losses and reputational damage.

To combat this surge, technology providers are working on improving fraud detection systems. Socure, for example, unveiled its graph intelligence module, following the company’s acquisition of fraud detection firm Effectiv. The technology, which uses machine learning to analyze connections between user behaviors, devices, and transaction histories, promises to detect fraud patterns earlier and more accurately than traditional methods.

The company claims that organizations adopting graph intelligence can achieve improvements in their fraud prevention efforts. The module provides fraud and data science teams with data, and offers a view of identity relationships, in a bid to reduce false positives.

Article Topics

BioCatch  |  biometrics  |  deepfakes  |  digital identity  |  financial crime  |  FinCEN  |  fraud prevention  |  identity verification  |  Sift  |  Socure