GCRI · Foundry
AI and agents
Foundry is the technical work environment of the Nexus Consortiums, and its public gateway is published by The Global Centre for Risk and Innovation (GCRI). AI agents and automated tools can be useful in that environment only if they act within limits that people set, record and can stop. This page explains which tool actions agents are designed to be permitted, how agents and models are evaluated before they receive authority, and the boundaries that apply to every agent action. Information and enquiries are open today.
AI and agents on the record.
- 01FieldsData and compute, Methods and modelling
- 02STEM methods6 canonical methods applied
- 03Topics8 subject pages
- 04Counterparts1 platforms in shared fields
- 05PublisherThe Global Centre for Risk and Innovation
An agent acts only within a scoped authority record.
In the design of the NXS-NSF (trust, authorisation and assurance), every agent and tool holds an authority record that states its purpose, the tools it may call, the data classes it may touch, its rate limits and its kill-switch. Requests without an explicit policy decision fail closed, including agent and tool calls.
Declared tools and data
An agent calls only the tools and reads only the data its record names, inside one authorised project.
The same receipts as people
Agent actions pass the same policy gates and receive the same receipts as human actions.
Kill-switches at the gate
Scopes, limits and kill-switches are enforced at the gate, not left to the agent’s own configuration.
Every action is bound to an identity and a supervisor.
In research work in NXS-LAB (research, experiments and simulations), agents may run analyses and draft artefacts, but each action is bound to an agent identity, its configuration, model, tool set, budget and supervising institution, and it remains an assertion until a person reviews it. In NXS-STU (composition, dashboards and workflows), agent behaviour inside a composition is declared with its purpose, tools and scope, and the trust spine enforces that scope at runtime.
Evaluation comes before authority.
- 01
Contract testing
Integrations and agents are tested against their declared contracts before they receive authority in any tenant.
- 02
Model evaluation
Promotion of a model requires evaluation records, including red-team results where the use calls for them.
- 03
Data use
No tenant data is used for training without a recorded decision by the data owner.
- 04
Review of outputs
Agent outputs used in evidence, reports or decisions are reviewed and attributed to the people who accept them.
What agents never do.
- Take a decision that belongs to an institution, a public servant or a competent authority.
- Issue a warning, release funds, dispatch resources or sign an acceptance.
- Act outside an authorised project, or on data the project has no right to use.
- Present a draft as reviewed evidence.
Public questions about responsible AI are taken up on GCRI responsible technology, and by The Global Risks Alliance (GRA) for financial services at GRA AI and models.
Questions and answers.
Can I connect my own agent to Foundry?
Who is responsible for what an agent does?
Can an agent be stopped immediately?
Is my data used to train models?
Bring a method, a dataset or a question.
Information and enquiries are open today. 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.
Legal documents.
Maps, data and figures on this site are shown for context and general information. They are not advice, forecasts or warnings; the limits of the information are set out in the Disclaimers and Data Notices.