Two individuals were the largest victims of pension fraud in England and Wales last year, each losing more than £400,000, according to data obtained from law enforcement. City of London police reported a total of £7.7 million lost across 242 pension fraud cases during the last financial year. Among these, 19 savers suffered individual losses exceeding £100,000.

Most fraud reports involved pension holders who had already accessed their funds, accounting for 166 cases, or 69 percent of the total. Criminals targeted this group through various schemes, including impersonation of bank or pension staff, false promises of high-return investments, and romance scams designed to divert pension payments.

The remaining 31 percent of reports, or 76 cases, involved fraudsters who convinced pension savers to pay upfront fees in exchange for early access to pension funds before the legally permitted age of 55. These deceptive scams frequently fail to disclose the tax consequences imposed by HM Revenue & Customs (HMRC).

Withdrawals made before age 55 are classified as unauthorized by HMRC and subject to a flat income tax charge of 40 percent, with an additional 15 percent surcharge if the withdrawal reaches 25 percent of the pension pot. When combined with scam fees, the financial penalty often results in the pensioner losing most or all of their funds.

Zurab Kotaria, chief executive of Identomath, a compliance firm that analysed the data, emphasized the need for improved security measures. He noted that traditional verification methods relying on security questions are no longer sufficient to prevent these frauds. Kotaria advocated for pension providers and banks to adopt biometric liveness scans that assess facial movements to verify identity, stating these could effectively thwart scammers.

The rise of artificial intelligence has further complicated efforts to combat pension fraud. Fraudsters are increasingly using AI tools to create convincing emails, mimic voices, replicate writing styles, and conduct deceptive video calls, making it harder for individuals and institutions to discern legitimate communications from fraudulent ones.