AI making financial crime ‘cheap, fast and invisible’, FCA warns

TL;DR:

  • The FCA’s joint enforcement director says AI is letting criminals innovate faster than law enforcement can respond.
  • Therese Chambers described the threat as “cheap, fast and invisible” against enforcement that is “expensive, slow and highly visible”.
  • The regulator is leaning on the “credible threat of enforcement” to intervene within weeks rather than years.

AI is accelerating the pace and scale of financial crime, the Financial Conduct Authority has warned, with criminals now innovating faster than enforcement can keep up. Speaking to the International Bar Association’s Anti-Corruption Conference, Therese Chambers, the FCA’s joint executive director of enforcement and market oversight, said AI has made the criminal threat “cheap, fast and invisible” while traditional enforcement remains “expensive, slow and highly visible”.

Closing the gap

Chambers argued the regulator must adapt to close that gap, increasingly using the “credible threat of enforcement” to act earlier — within weeks rather than months or years. She said FCA investigations are moving faster than ever, with 10 reaching a public outcome in 16 months or less since July 2024. The watchdog recorded 369 voluntary outcomes last year and supported 124 more complex cases through its interventions team, against 13 cases requiring formal powers.

The warning fits a worsening UK fraud picture, from the surge in AI-enabled romance scams to mounting concern that financial firms are sleepwalking into AI governance failures. It also lands as the ICO sets out its plan for safe AI innovation, underscoring how UK regulators are racing to match the technology criminals are already exploiting.

Looking forward

Chambers was clear that much of the threat “does not originate on our shores”, a reminder that AI-enabled fraud is a cross-border problem no single regulator can solve alone. For UK financial firms, the practical takeaway is that the FCA intends to move at speed — and expects the institutions it oversees to detect and disrupt AI-driven crime just as quickly. The harder question is whether a regulator built for deliberate, evidence-heavy enforcement can genuinely operate at the tempo synthetic-media and automated fraud now demand.

Share this article