Banks focus on catastrophic AI risks while AI-powered fraud already drains accounts one transaction at a time

Regulators (like those at the Mythos meeting) worry AI could crash financial systems or manipulate markets at scale. Meanwhile, fraudsters use AI to impersonate customers, bypass authentication, and authorize unauthorized transfers — losses are already happening but scattered enough to stay under the radar.

AI-enabled fraud bypasses traditional detection because it mimics legitimate customer behavior (the speed and pattern of real transactions) rather than triggering alarms like a sudden $1 million wire transfer would. This means banks' existing fraud filters — built to catch obvious outliers — don't catch it.

For you: If you bank online, your account is already a target for AI-driven impersonation attacks. Your bank's fraud team is focused on preventing system-wide crashes, not the incremental theft from thousands of individual accounts. This gap means personal account security now depends more on your own defenses (strong passwords, authentication apps) than your bank's institutional ones.