HomeEverythingEducationTV
Equities & FundsCrypto & Digital AssetsAI & TechnologyBusiness & CorporateUS Politics & PolicyGeopolitics & Global RiskMacro, Rates & FXCommodities & EnergyEuropean Politics & MarketsAsia-PacificReal Estate & Property
Story archiveAll categories
← All Stories

Banks Lack Failure Plans for Tier 1 Models

Created at 22 Jul · 3:36 AM1 source↑ Market-relevant
IN SHORT

A Risk.net study reveals inconsistent control frameworks for model failure post-deployment among banks, even for critical Tier 1 models. This includes a lack of documented plans for recognizing and responding to breaches.

✉Newsletter

PiQ Daily

Pick your topics. Get only what matters, on your cadence.

Who's Involved

Risk.net
Publisher of the Model Risk Benchmarking study

↳ Why This Matters

The lack of robust failure plans for critical models poses a significant risk to financial institutions, potentially leading to undetected breaches and inadequate responses, which could have severe financial and reputational consequences.

Key facts

  • Control frameworks for recognizing and responding to model failures post-deployment are inconsistent among banks.
  • This inconsistency extends to critical Tier 1 models.
  • Many banks lack documented plans for handling model breaches.
  • The findings are part of Risk.net's Model Risk Benchmarking study.

A study by Risk.net has found that banks exhibit inconsistencies in their control frameworks for managing model failures after deployment, even for their most critical Tier 1 models. The research indicates a significant gap in documented plans for identifying and addressing breaches. This issue is part of an ongoing series by Risk.net that benchmarks bank model risk management practices. The study highlights that while predeployment validation may be strong, the escalation of breaches post-deployment is often handled in an ad hoc manner, affecting even large financial institutions.

Frequently asked questions

Tier 1 models are typically considered the most critical models used by a bank, often having the most significant impact on financial decisions, risk assessment, or regulatory compliance.

Model failure refers to instances where a deployed model no longer performs as intended, produces inaccurate results, or fails to meet its objectives, potentially due to data drift, changes in the environment, or inherent flaws.

It is a series of studies conducted by Risk.net to assess and compare the model risk management practices of various financial institutions.

What Happens Next

01Banks are expected to improve their model risk management practices.
02Further data cuts from the study will be available to subscribers.

Get the newsletter.

Pick the topics you actually care about. We'll email when there's news worth your time, on the cadence you choose. Cancel any time from your account.

Cadence

How It Developed

Control frameworks for model failure post-deployment are inconsistent among banks.
This inconsistency affects even critical Tier 1 models.
Banks lack documented plans for recognizing and responding to model breaches.

Sources

T1
Many banks do not document failure plans for Tier 1 modelsRisk.net

Related Stories

OpenAI AI Models Escaped Test Environment to Hack Hugging Face
21 Jul · 9:31 PM
UK datacentre growth plans face water supply challenges, trade body warns
21 Jul · 4:11 AM
Social platforms update AI content rules amid concerns over spam
22 Jul · 12:36 AM
AI hackers thwarted by 'banned topics' in security tests
21 Jul · 7:11 AM
Bank of America enhances AI tool to speed client service
21 Jul · 1:16 PM