Banks refine scam controls as AI attacks rise

Outseer presented new consumer survey data in a webinar that examined which bank controls most effectively limit consumer harm as AI-powered and “scam-as-a-service” attacks increase. The survey covered more than 15 markets.

Panelists noted organised scam services and AI tools are lowering barriers for attackers and cited an industry projection that losses for financial institutions could reach USD 55.3 billion by 2030.

Speakers used a five-stage scam framework-pre-attack, consumer compromise, transaction, receipt and post-attack-to review controls. For each stage they discussed early detection and blocking, ways to reduce exposure to social engineering, transaction monitoring and real-time warnings, verification at receipt, and options for remediation and recovery.

Panelists identified consumer behaviour as a primary target of modern scams, with attackers increasingly exploiting human vulnerability rather than technical gaps. That perspective framed questions about how far technical controls can be relied on and where behavioural science should be applied.

Survey findings presented by Outseer included data on the controls consumers value and trust. Panelists reported that customers expect stronger protection, clearer communication and faster support when targeted. The webinar summary did not publish detailed regional breakdowns.

Speakers included Martyn Higson, principal product manager at Outseer. The session was moderated by Sharon Kimathi. Other industry experts discussed strategies used by banks and emerging approaches intended to respond to more automated and outsourced scam operations.

In closing, panelists recommended combining technical measures with behavioural interventions and prioritising controls that intervene earlier in the scam lifecycle to reduce customer harm and institutional losses.