The Global Risks Alliance (GRA)
The methods AI and Financial Model Risk relies on, the assumptions they carry and the limits of what they can show.
Fig.How a finding is reviewed before it is published. Design, not a record.
The AI and Financial Model Risk methods page sets out the methods the platform relies on, among them artificial intelligence and machine learning, AI evaluation and assurance engineering, and uncertainty quantification and model validation, with their assumptions, validation and limits. Each analysis lists its assumptions and data, reports ranges and confidence beside central estimates, is checked against observations or other models, and states what it cannot show.
For analysis and uncertainty, in GRA's spatial-finance reading, the map opens on night lights and population density with night lights tonight from NASA GIBS, framed on California; 4 further NASA layers wait in the list.
The interactive map needs JavaScript; the layer list and sources are below.
Layers 2
Connected activity, for this page
Keyboard, with the map focused
Hold Ctrl or ⌘ and scroll to zoom
Latest available
NASA GIBS / EOSDIS · ASTER GDEM (METI, NASA) · Blue Marble (NASA) · © OpenStreetMap contributors (ODbL) · Natural Earth
Loading the latest NASA imagery.
Drag to move. Ctrl or ⌘ and scroll, pinch or the buttons to zoom; hover for the values under the pointer; reset returns to the region.
For analysis and uncertainty, in GRA's spatial-finance reading, the map opens on night lights and population density with night lights tonight from NASA GIBS, framed on California; 4 further NASA layers wait in the list.Satellite and model products as NASA publishes them, each with its own date; indicators, not assessments of any place; claim ceiling INTEGRATED.
Layers and dates
| Method | Gives | Limit |
|---|---|---|
| SHAP and other attribution methods | Contribution of each input to a prediction | Correlated inputs blur attribution |
| Counterfactual explanations | What change would alter the decision | May suggest infeasible changes |
| Group fairness metrics | Outcome differences across groups | Metrics can conflict |
| Challenger models | Independent benchmark | Shares data weaknesses |
Prediction intervals and calibration checks show when a model is outside its experience. Methods at GCRI uncertainty quantification; topic depth at explainability and fairness.
Each analysis lists its assumptions and the data it depends on.
Ranges and confidence are reported alongside central estimates.
Models are checked against observations and, where possible, against other models.
What an analysis cannot show is stated beside what it can.
Each term as this page uses it. Where a standard-setting body defines the term, the wording follows that body.
| Term | Definition |
|---|---|
| Uncertainty quantification | Characterising and propagating the uncertainties in a model to state confidence in its outputs. |
| Sensitivity analysis | Repeating an analysis under different assumptions to show how much the result depends on them. |
| Model validation | Independent testing that a model performs as intended and is fit for a stated use. |
| Artificial intelligence and machine learning | Artificial intelligence and machine learning develop algorithms that learn patterns from data to classify, predict or generate outputs. |
The bodies and documents this page is read against. Status and editions change; each link goes to the issuing body.
The current service level is R0: information and enquiries. An enquiry receives a reference and names the next step; applications, payments, compute allocations and registrations open only when their team, terms and review capacity exist.