Monzo Bank fined £21 million by FCA for AML control failures
Monzo Bank received a £21 million penalty from the FCA due to inadequate anti-money laundering controls. The failure was characterized by a lack of scalable automation and systemic gaps in customer due diligence.
Monzo's automated monitoring failed to scale, leaving staff unable to identify the specific transactions that triggered alerts.
Key facts
- What
- Monzo Bank received a £21 million penalty from the FCA due to inadequate anti-money laundering controls.
- Incident date
- Jul 7, 2025
- Who
- Monzo Bank
- Failure mode
- Hallucination
- AI surface
- Search / RAG
- Severity
- High
What happened
The Financial Conduct Authority fined Monzo £21 million for systemic failures in its anti-money laundering systems and controls. The bank failed to perform adequate customer due diligence and risk assessments between 2018 and 2020. Additionally, Monzo violated a regulatory ban by onboarding over 34,000 high-risk customers between 2020 and 2022.
What broke inside the model
- 01 · TriggerA user asks for a fact, a citation, or a figure.
- 02 · Model stepThe model writes a fluent, confident answer.
- 03 · Control gapNothing ties the claim back to a real source.
- 04 · FailureA fabricated fact ships as if it were verified.
- 05 · ConsequenceThe false claim reaches a customer, a court, or the public.
Confidence holds, and even spikes, as the claim detaches from any source.
The bank's automated monitoring systems failed to scale with its rapid growth, creating an opaque environment where alerts were generated without clear pointers to the triggering transactions. This lack of transparency prevented staff from effectively identifying and investigating suspicious activity.
What it cost
Sources
Cite this entry
https://failureindex.ai/failures/monzo-bank-fined-million-fca-amlAI Failure Index. "Monzo Bank fined £21 million by FCA for AML control failures" (FI-0492). Realm Labs. https://failureindex.ai/failures/monzo-bank-fined-million-fca-aml (indexed Jun 10, 2026).Data fields CC-BY 4.0, prose citation permitted. Incident ID FI-0492. Full dataset at /data.
Note from Realm Labs, the Index steward
How Realm would have caught this
- Prism
- OmniGuard
- AI Detection & Response (AIDR)
A runtime layer that watches the model's internal state can flag the moment a model commits to a claim it has no support for, and hold or reroute the response before it reaches a user. Realm reads those signals in real time rather than grading the transcript after the fact.