Risk Management
Latest Risk Management
The do-it-all machine: model risk in the age of generative AI
As LLMs take on more tasks, banks are under pressure to manage the risks from hundreds if not thousands of new applications.
Of 13 banks that Risk.net canvassed, only Bank of America and Goldman Sachs subject LLMs to model validation by default. The remainder, including Citi and JP Morgan, take a flexible approach depending on the level of risk.
The divergence is fuelling a debate over whether the SR 11-7 supervisory guidance for model risk is still fit for purpose.
In the age of GenAI, why do we still need good models?
Jean-Philippe Bouchaud says models can guide artificial intelligence through regime shifts and away from overfitting
Top 10 op risks: AI upends risk taxonomies
AI risk enters annual poll in fifth, but firms split over treating it as a standalone risk or a cross-cutting driver
EBA proposes drastic overhaul to supervisory data reporting
Revamp will cut back the number of datapoints and integrate overlapping reports
CFTC wants to regulate prediction markets. Is it up to the task?
Former officials echo state gambling authorities’ concerns over agency’s ability to police betting risks
Mob rule: populism’s rise pits banks against the people
Trump and fellow mavericks are reshaping politics, leaving banks scrambling to adjust to new and unpredictable risks
Climate risk
Most banks run physical climate scenarios beyond 2050
Risk Benchmarking data finds majority rely on geospatial asset mapping, while a third use third-party catastrophe models
Big banks love their climate vendors; small banks, not so much
Risk Benchmarking: Lenders with blue-chip loan books more likely to favour climate tools, research finds
Climate risk managers’ top challenge: a dearth of data
Risk Benchmarking: Banks see client engagement and lender data pooling as solutions to climate blind spots – but few expect it to happen soon
Why better climate data doesn’t always mean better decision-making
Risk Benchmarking research finds model and systems integration challenges almost as limiting to effective climate risk management
Risk Quantum
Data insights, delivered daily
Risk Quantum finds insights in data. The service tracks the public disclosures of over 120 banks, funds, insurers, corporates, and central counterparties – as well as reports from prudential and markets regulators – in Asia, Europe and North America.
Wells Fargo tops US banks’ financing of private credit
Bank reports $71bn in loans and commitments, but reporting differences cloud comparisons
Artificial intelligence
FHLB Cincinnati explores AI to spot failing banks
Agentic model detects anomalies, monitors sentiment and drafts credit reports for analyst review
The MIT professor giving LLMs a ‘brain scan’
Hui Chen’s research is yielding new ways to interpret – and steer – AI models
At BNY, a risk-centric approach to GenAI
Centralised platform allows bank to focus on risk management, governance and, not least, talent in its AI build
Banks split over AI risk management
Model teams hold the reins, but some argue AI is an enterprise risk
We’re gonna need a bigger board: geopolitical risk takes centre stage
As threats multiply, responsibility for geopolitical risk is shifting to ERM teams
Sponsored content
About
These articles were paid for by contributing third parties. Click here for more information on content funding.
