AI-Driven Fraud, Collective Defence & the Exponential AI Curve | May ’26 AI Bulletin

The Challenge

AI-driven fraud is scaling at a pace the sector hasn’t seen before. Fraud losses through the banking system exceeded £1bn in 2025. Voice cloning takes 20 seconds of audio. Deepfakes are bypassing identity checks that lenders have relied on for years. The tools are getting cheaper and more accessible every month.

Why Collective Defence and HDI Matter Now

No single lender can tackle this alone. Across the sector, particularly among mutuals, there is a growing appetite to pool resources, combine expertise and build shared defences. Industry-wide collaboration and HDI to operationalise broad fraud protection is where the sector needs to move.

The Same Capabilities: Threat and Opportunity

The same AI that enables faster document processing, identity verification and anomaly detection is what criminals are using. Building AI capability and defending against AI threats are not separate investments. Governed data flows are harder to compromise. Structured digital identity is harder to fake.

HDI builds resilience at the same time as capability.

When data flows through verified, governed rails end to end, fraud finds fewer places to enter. Fragmented, manual processes are where the risk sits.

Three Practical Steps for Lenders

Review your identity verification against deepfake capability. Engage with collective defence across the sector rather than building walls alone. Get on the AI curve now and consider your enterprise governance layer. AI capability is exponential, and the gap is widening every month.