Skip to content

GCRI · Foundations

Contribution without capture: how sponsors, vendors and technology providers take part in the Nexus programme

Sponsors, vendors and technology providers can add money, computing, data and people to the Nexus programme, but none of it buys a finding, a seat or a decision. This page sets out the rules, the records that make them checkable, and the evidence for why they matter.

As of
21 October 2026
Reading
13 minutes
Evidence
13 sources, cited in order
NEXUS TWIN · WF-1 · NETWORKSEED bdde30eeGCRI-MAIN.POST.1076862xyzSCHEMATIC · NETWORKST-01 · 0.762 0.443ST-02 · 0.359 0.463ST-03 · 0.857 0.358
WF-1 · network Schematic wireframe twin (network, decision under uncertainty). An illustration, not data.

Methods and evidence

Evidence
13 sources, cited in order and listed under References
Figures
4 figures
Data availability
Every chart's data is published as a table on this page.
Bias
1.27
risk ratio for favourable efficacy results in industry-sponsored studies3
States
7
recorded states between an offer of capacity and its acceptance4
Price list
None
no fee or contribution amount buys hosting or sponsorship2
Service
R0
information and enquiries; calls open with their terms5

01What contribution without capture means

What contribution without capture means

Contribution without capture is the principle that any organization may contribute money, technology, computing, data, people or venues to the public-good work of the Nexus Consortium while acquiring no influence over the evidence, the agenda, the selection of participants, publication or any decision. The Global Centre for Risk and Innovation (GCRI) states it in three words on its independence and conflicts page, "support buys nothing", and pairs it with disclosure of funding, declaration of interests and selection on documented criteria 1. The Global Risks Forum (GRF) states the same rule for hosts and supporters of the annual programme: hosting is not ownership, and sponsorship is not authority 2. The Global Risks Alliance (GRA) states it for finance: funding buys no control 6.

This page replaces an earlier essay on the same subject. It answers the question a sponsor, vendor, technology provider or public official will ask first: if an organization contributes, what does it receive, what does it never receive, and how would anyone know if the rule were broken?

02Why capture is the central risk of shared evidence

Why capture is the central risk of shared evidence

Shared evidence work creates a risk of capture. The OECD defines policy capture as "the process of consistently or repeatedly directing public policy decisions away from the public interest towards the interests of a specific interest group or person", achieved through illegal means such as bribery but also through legal channels such as lobbying and financial support, and even by manipulating the information decision makers rely on (Preventing Policy Capture) 7.

The strongest empirical evidence comes from medicine, where sponsorship has been studied longest. A Cochrane methodology review of 75 papers found that drug and device studies sponsored by the manufacturer more often had favourable efficacy results (risk ratio 1.27, 95% confidence interval 1.17 to 1.37) and favourable conclusions (1.34, 1.19 to 1.51) than studies with other sponsors, and that their conclusions agreed less often with their own results (0.83, 0.70 to 0.98) 3. The authors concluded that the pattern suggests "an industry bias that cannot be explained by standard 'Risk of bias' assessments" 3. Nothing in that finding depends on bad faith. Bias enters through choice of questions, comparators, framing and emphasis, which is exactly why a rule against capture has to govern the whole chain from agenda to publication, not only the final report.

For systemic risk the stakes are higher than in a single trial. A scenario designed around one vendor's platform, or a stress test that softens the exposure of a sponsor, can shift what governments, operators and financial institutions believe about their own vulnerability. GRF names the risk directly for its exercises: a tested host, or a supporter with a stake in a scenario, could soften what is found 2.

Fig. 01Sponsorship shifts results, measurably

Sponsorship shifts results, measurably Forest plot of five pooled risk ratios from a Cochrane review of 75 papers: favourable efficacy results 1.27, favourable conclusions 1.34, similar harms 1.37 with a wide interval crossing 1, low risk of bias from blinding 1.25, and agreement of conclusions with results 0.83. EstimateIntervalrisk ratio0.1110No differenceFavourable efficacy1.27 (1.17 to 1.37)Favourable conclusions1.34 (1.19 to 1.51)Harms1.37 (0.64 to 2.93)Blinding low risk1.25 (1.05 to 1.5)Results agree with conclusions0.83 (0.7 to 0.98)
A risk ratio above 1 means the outcome was more frequent in industry-sponsored studies; intervals that cross 1 are not statistically distinguishable from no difference. Log scale.3Sources: Lundh et al., Cochrane Database of Systematic Reviews (2017). As of 21 October 2026.
Figure 1 in words

A forest plot on a logarithmic axis from 0.5 to 3. Favourable efficacy results: risk ratio 1.27, 95% confidence interval 1.17 to 1.37, 25 papers, moderate-quality evidence. Favourable conclusions: 1.34, 1.19 to 1.51, 29 papers, low-quality evidence. Harms results: 1.37, 0.64 to 2.93, four papers, very low quality, the interval crosses 1. Low risk of bias from blinding: 1.25, 1.05 to 1.50, 13 papers. Agreement between results and conclusions: 0.83, 0.70 to 0.98, six papers, meaning less agreement in industry-sponsored studies. A vertical line marks a ratio of 1, no difference.

03What the instruments of public integrity require

What the instruments of public integrity require

The programme's rule follows the instruments that intergovernmental bodies already apply to themselves. The OECD Recommendation on Guidelines for Managing Conflict of Interest in the Public Service, adopted on 28 May 2003, sets four principles: serving the public interest, supporting transparency and scrutiny, promoting individual responsibility and personal example, and engendering an organisational culture that is intolerant of conflicts of interest (OECD) 8. The World Health Organization reviews every engagement with non-State actors, including financing, in-kind contributions and secondments, under its Framework of Engagement with Non-State Actors, adopted by the Sixty-ninth World Health Assembly in resolution WHA69.10 (WHO) 9.

The Intergovernmental Panel on Climate Change applies a conflict-of-interest policy to its leadership, authors, review editors and technical support staff, so that conflicts are "identified, communicated to the relevant parties, and managed to avoid any adverse impact on IPCC balance, products and processes" (IPCC) 10. The policy notes that identifying a potential conflict does not automatically mean that one exists; its purpose is to put the facts on record so that each case can be judged 10. Three lessons carry over to the Nexus programme: declare before taking part, record what was contributed and on what terms, and keep the people who decide content apart from the people who fund it.

YearInstrumentWhat it requires
2003OECD Guidelines for Managing Conflict of Interest in the Public ServiceFour principles for managing conflicts of interest, from serving the public interest to an organisational culture intolerant of them 8
2011, amended 2016IPCC Conflict of Interest PolicyConflicts identified, communicated and managed for leaders, authors, review editors and staff 10
2016WHO Framework of Engagement with Non-State Actors (WHA69.10)Review of financing, in-kind contributions and secondments from non-State actors 9
2017OECD, Preventing Policy CaptureA framework for stakeholder engagement, transparency and accountability 7
Instruments of public integrity that the programme's contribution rules follow, with their status on 21 October 2026.

04Who contributes, and through which door

Who contributes, and through which door

The programme describes every external party by role, never by name, and gives each role its own route, first outcome and boundary 11.

  • SMEs, startups and technology teams respond to a testbed or mission call. The first outcome is a documented selection decision, an evaluation plan and controlled results; evaluation never depends on contribution size, and testing is not certification 4,11.
  • Computing centres, research networks, telecoms, cloud and equipment providers offer qualified resources or technical capacity. The first outcome is a resource-offer disposition and a qualification record; allocations are approved separately, and offered capacity is never counted as reserved, consumed or funded 4,11. The annual computing effort is described on network and core.
  • Enterprises, manufacturers and industry associations explore a sector programme with a work scope, permitted data exchange and evaluation terms; precompetitive work never shares competitors' confidential information 11. Operators start from infrastructure and industry.
  • Sponsors and supporters fund a defined public-good capability under a contribution agreement with a use-of-resources report; hosts provide a venue or operational support under a hosting agreement with a responsibility schedule and a closeout record 2. GRF sets the terms in its sponsorship policies 12.
  • Standards bodies and professional societies propose methods or interoperability collaboration in the open standards process; test completion is never presented as accredited certification 11. The route is develop a standard.

Fig. 02Five doors, one boundary

Five doors, one boundary Five columns for technology teams, capacity providers, industry, sponsors and hosts, and standards bodies, each with its first step, its first outcome and its boundary, such as evaluation never depending on contribution size and offered capacity never counted as reserved. First stepFirst outcomeBoundaryStrong linkPartial link
Routes as published in the programme's stakeholder model; every route starts with an enquiry at service level R0.4,2,11Sources: GCRI technology and infrastructure providers; GRF hosts and supporters; programme documentation. As of 21 October 2026.
Figure 2 in words

Column one, SMEs, startups and technology teams: respond to a testbed or mission call; first outcome a documented selection decision, evaluation plan and controlled results; boundary, evaluation never depends on contribution size and testing is not certification. Column two, computing centres, networks, telecoms, cloud and equipment providers: offer qualified capacity; first outcome a resource-offer disposition and qualification record; boundary, offered capacity is never counted as reserved, consumed or funded. Column three, enterprises and industry associations: explore a sector programme; first outcome a work scope, permitted data exchange and evaluation terms; boundary, precompetitive work never shares competitors' confidential information. Column four, sponsors and hosts: a contribution or hosting agreement; first outcome a use-of-resources report or a responsibility schedule; boundary, hosting is not ownership and sponsorship is not authority. Column five, standards bodies and professional societies: propose methods; first outcome a scoped profile; boundary, test completion is never presented as accredited certification.

05How a contribution is recorded, state by state

How a contribution is recorded, state by state

The core safeguard is that a contribution never jumps from offer to influence. GCRI records an offer of capacity with its owner and contact, legal and use constraints, location and availability window, hardware and software profile, security state, support coverage and its return and sanitization requirements. The offer then moves through separate states, offered, qualified, reserved, allocated, accessible, measured and accepted, and none of them implies the next 4. Commitment arises only through a separately approved allocation with agreed terms 4.

The same logic governs the programme as a whole. Its register of separations states that an enquiry is not membership, an appointment or a service contract; a payment is not an appointment or access to restricted data; an infrastructure offer is not an allocation; and readiness evidence is not a financing or underwriting decision 5. For provider agreements, the accountable GCRI organization is the one named in the terms, GCRI Canada or GCRI US 4.

In the shared evidence environment, three systems carry these records. NXS-RLS (participants, agreements and handoffs) holds who contributed what, under which agreement and with which handoff; NXS-COR (compute, nodes and runtimes) holds the qualification and measurement of contributed computing; and NXS-NSF (trust, authorization and assurance) governs access, so that contributing a resource never confers access to restricted data. The Nexus Standards profile for system and stack passports, NRS-S0, is the proposed common form for describing what a contributed component is and how it was tested.

Fig. 03An offer is not an allocation

An offer is not an allocation A seven-step ladder for contributed capacity: offered, qualified, reserved, allocated, accessible, measured and accepted, with a gate before allocation marking where separate approval and agreed terms are needed, and a panel listing what an offer must record.
Each state has its own record and approval. Commitment arises only at allocation, under separately agreed terms.4,11Sources: GCRI technology and infrastructure providers; programme documentation. As of 21 October 2026.
Figure 3 in words

Seven steps from left to right. Offered: the owner records the resource with contact, legal and use constraints, location and availability window, hardware and software profile, security state, support coverage and return and sanitization requirements. Qualified: the resource is checked against the mission's needs. Reserved: it is held for a mission. Allocated: a separately approved commitment with agreed terms, marked as the point of commitment. Accessible: authorised users can reach it. Measured: use and performance are recorded. Accepted: the contribution is accepted into the record. A banner states that none of these states implies the next, and that an offer is never counted as reserved, consumed or funded.

06What sponsorship and hosting can and cannot buy

What sponsorship and hosting can and cannot buy

GRF's rules for the annual programme, Nexus Universe, are the most detailed statement of the boundary, because an exercise is more exposed than a conference. They list what support never buys: no support buys an appointment, a finding, consent or a procurement advantage; sponsors cannot approve publications, change evidence or choose participants; funding grants no ownership of the programme, its records or its results; neither hosting nor sponsorship secures a computing allocation or mission admission; and a host or sponsor is named only with its written permission and a current record 2.

The rules also match each form of support to the specific risk it carries and to a control designed for that risk. They are reproduced below because the matching, rather than a general promise of independence, is what makes the boundary testable.

Form of supportRisk to independenceControl
Venue or in-kind supportHost expects visibility over contentHost has no role in scenario or record 2
Funding for a capabilityFunder expects the capability to favour its interestsCapability defined before funding, with a public description 2
Secondment of staffSeconded staff carry employer prioritiesDeclared, and excluded from evaluating their employer 2
Data contributionData shapes the scenarioProvenance and licence recorded; scenario reviewed by others 2
Forms of support, the risk each carries to independence, and the control applied, as published by GRF for Nexus Universe on 21 October 2026.

07Claims a contributor may and may not make

Claims a contributor may and may not make

Capture often arrives after the work, through what is said about it. GCRI's provider terms answer the question directly: participation may be used in marketing only as an accurate statement of what took place, with published results cited in full, and may not be described as endorsement, partnership or certification 4. A test result describes what was tested; it is not accreditation, certification, endorsement or exclusivity 4. Exclusivity is not granted by default, and technology providers compete on evidence 1.

Disclosure runs the other way too. Supporters and the capability they fund are listed in the edition record with the form of support and any conditions, and anonymous support is not accepted 2. No fee or contribution amount is published as a price for hosting or sponsorship; any contribution is set in written terms with the issuer, currency, effective date and scope 2. GRA applies the same sequence to its own members and sponsors, declare, record, recuse and disclose, and explains on membership and authority why no member's interest is advanced through its work 6.

Fig. 04What participation is, and what it is not

What participation is, and what it is not A two-column comparison of eight pairs: a test result is not certification; an offer is not an allocation; sponsorship is not authority; hosting is not ownership; a payment is not an appointment or data access; a secondment is not an evaluator role; participation is not endorsement; readiness evidence is not a financing decision.
Each line pairs a step with the status it never confers. The pairs are published by GCRI and GRF and govern any page that claims more.4,2,5Sources: GCRI providers and claims register; GRF hosts and supporters. As of 21 October 2026.
Figure 4 in words

Left column, the step; right column, what it is not. A test result is not accreditation, certification, endorsement or exclusivity. An infrastructure offer is not an allocation. Sponsorship is not authority over agenda, participants, publications or findings. Hosting is not ownership of the programme, its records or results. A payment is not an appointment or access to restricted data. A secondment is not a role in evaluating the employer. Participation in a test or exercise is not endorsement or partnership. Readiness evidence is not a financing or underwriting decision.

08How the standards process keeps providers at arm's length

How the standards process keeps providers at arm's length

The Nexus Standards are proposed, not adopted, and their governance is designed so that no provider can shape them from the inside. The twenty systems known by their NXS codes are proprietary infrastructure, not open and not free, and no standard depends on them; an external provider supports them under written agreements and holds no vote, chair or Secretariat role in the governance of the standards 11. Background intellectual property stays with its owner, and each test sets out the licence for provider technology and for results 4.

09How anyone can check that the rule holds

How anyone can check that the rule holds

A rule against capture is credible only if an outsider can test it. Three records make that possible. The contribution record states what was contributed, by whom, under what terms and with what limitations. The edition record of each programme year lists supporters and the capability each funded 2. And the claims register shows what the programme has planned, tested and provided, and it governs any page that claims more 5.

Anyone, member or not, can raise a concern. Concerns about a provider relationship go to the integrity concern form, and a scientific correction or integrity concern is open to anyone without membership and is reviewed by the responsible editor or source owner 4. A complaint about a programme decision is reviewed by someone who was not responsible for the decision being contested. GRF takes concerns about capture through institutional independence 2. Findings published in Reports carry their sources, limits and corrections, so a reader can see whether a contributed dataset or model shaped a result.

10What is changing

What is changing

  • Bias survives standard checks. The Cochrane review found an industry bias that standard risk-of-bias assessments do not explain, which is a reason to separate who funds from who decides, and not only to review more carefully 3.

11Three institutions, one boundary

Three institutions, one boundary

Each institution applies the rule where its exposure lies. GCRI applies it to technology, testbeds and contributed capacity, where the risk is that a test becomes an endorsement or an offer becomes a claim on the programme 4,1. GRF applies it to convening, hosting and the annual exercise, where the risk is that a host or supporter shapes a scenario or an agenda 2,12. GRA applies it to finance, where it recommends no investor, lender, insurer, vendor or project and takes no pay tied to any transaction 6. The decision rights of the programme put the final decision with the competent government, operator or licensed institution in every case.

12How to contribute now

How to contribute now

At service level R0, information and enquiries, every route below starts with a written enquiry, acknowledged with a reference and the name of the next step. Calls are published when they open, with their terms 4,5.

What this page does not claim: no sponsor, vendor, host or provider is named here or endorsed, and naming an instrument or organization is not a claim of partnership. A test result is not certification or accreditation, an offer is not an allocation, and a contribution confers no seat, vote, finding, data access or procurement status. No call, allocation or host is open or confirmed; the programme is at service level R0. GCRI does not regulate or certify, GRF does not regulate, finance or procure, and GRA does not lend, underwrite, rate or invest.

Questions

Questions and answers

What does a sponsor receive in return for support?

A contribution agreement for a defined public-good capability and a report on how the resources were used. A sponsor receives no influence over the agenda, participants, publications or findings, and is named only with written permission.

Can a vendor say its product was tested or endorsed in the programme?

A vendor may state accurately what took place, citing published results in full. It may not describe participation as endorsement, partnership or certification, because a test result describes only what was tested.

Does offering computing or data give a provider a place in the programme?

No. An offer is the first of seven recorded states. Commitment arises only through a separately approved allocation with agreed terms, and contributing a resource never confers access to restricted data.

Does a larger contribution improve the chance of being selected for a testbed?

No. Participants are selected on documented criteria, and evaluation never depends on the size of a contribution. Selection decisions are documented so that they can be checked.

How do I report a suspected conflict or capture?

Use the integrity concern form, or the GRF independence route for the annual programme. Concerns are open to anyone, with no membership needed, and complaints are reviewed by someone not responsible for the contested decision.

Is there a price list for sponsorship or hosting?

No. No fee or contribution amount is published. Any contribution is set in written terms with the issuer, currency, effective date and scope, and anonymous support is not accepted.

Terms

Glossary

Capture
The consistent or repeated direction of public decisions away from the public interest towards a narrow interest, by legal or illegal means.
Contribution record
The record of what was contributed, by whom, under what terms, with what limits and with which claims permitted.
Resource state
One of seven recorded states of contributed capacity, from offered to accepted; no state implies the next.
Allocation
A separately approved commitment of a qualified resource to a mission, with agreed terms.
Arm's length
A relationship under written agreement in which a provider holds no role in stewardship and no control over programme decisions.

Reuse and cite

Text CC BY 4.0. Cite the page with its date; data tables are linked under each figure.

Sources

References

13 sources, numbered in the order they are first cited
  1. GCRI, Trust and accountability (2026-10). Independence and conflicts. therisk.global. Accessed 2026-10-21.
  2. The Global Risks Forum (2026-10). Hosts and supporters. globalriskforum.com. Accessed 2026-10-21.
  3. Lundh, Lexchin, Mintzes, Schroll and Bero (2017-02-16). Industry sponsorship and research outcome (Cochrane Database of Systematic Reviews, MR000033) (Europe PMC record). www.ebi.ac.uk. Accessed 2026-10-21.
  4. The Global Centre for Risk and Innovation (2026-10). Technology and infrastructure providers. therisk.global. Accessed 2026-10-21.
  5. GCRI, Trust and accountability (2026-10). Claims and readiness register. therisk.global. Accessed 2026-10-21.
  6. The Global Risks Alliance (2026-10). Conflicts, sponsors and neutrality. globalriskalliance.com. Accessed 2026-10-21.
  7. OECD (2017-03). Preventing Policy Capture: Integrity in Public Decision Making. www.oecd.org. Accessed 2026-10-21.
  8. OECD (2003-05-28). Recommendation of the Council on OECD Guidelines for Managing Conflict of Interest in the Public Service (OECD/LEGAL/0316). legalinstruments.oecd.org. Accessed 2026-11-17.
  9. World Health Organization (2016-05). WHO's engagement with non-State actors: Framework of Engagement with Non-State Actors (FENSA, resolution WHA69.10). www.who.int. Accessed 2026-10-21.
  10. Intergovernmental Panel on Climate Change (2016-10). IPCC Conflict of Interest Policy (as amended, Decision IPCC/XLIV-3). www.ipcc.ch. Accessed 2026-10-21.
  11. Nexus programme documentation (GCRI) (2026-10). Technology and service providers (audience guide: routes, offerings and boundaries). docs.therisk.global. Accessed 2026-10-21.
  12. GRF, Participate (2026-10). Sponsorship policies. globalriskforum.com. Accessed 2026-10-21.
  13. The Global Centre for Risk and Innovation (2026-10). Experts, volunteers and students. therisk.global. Accessed 2026-10-21.

Record

Editorial record

Facts and programme states are given as of 21 October 2026. This page is reviewed at least every 180 days and whenever a contribution rule, a programme state or a cited figure changes. It replaces an earlier version of this post, published in June 2026, whose description of contribution intake, sponsor and provider record types and named infrastructure did not match the programme's current records; those descriptions are withdrawn and replaced by the published rules cited here.

Published by the GCRI editorial desk under the editorial integrity policy. To report an error or request a correction, use the corrections route; corrections are dated and listed with the page.

Cite this article

GCRI editorial desk (2026, June 8). Contribution without capture: sponsors, vendors, providers. The Global Centre for Risk and Innovation. https://therisk.global/sponsor-vendor-and-provider-participation-contribution-without-capture-in-the-nexus-ecosystem/

Republish

Republish or translate this article free of charge, in print or online, under CC BY 4.0: keep the byline and The Global Centre for Risk and Innovation, link to the original and note any change. Figures, maps and twins made by the institutions carry the same licence; third-party data keeps its own credit line.

Attribution to paste
<p>Contribution without capture: how sponsors, vendors and technology providers take part in the Nexus programme — GCRI editorial desk, The Global Centre for Risk and Innovation. First published 8 June 2026 at <a href="https://therisk.global/sponsor-vendor-and-provider-participation-contribution-without-capture-in-the-nexus-ecosystem/">https://therisk.global/sponsor-vendor-and-provider-participation-contribution-without-capture-in-the-nexus-ecosystem/</a>. Republished under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).</p>

Topics

Related across the Consortium

Share

EmailLinkedInBlueskyMastodon

The Global Centre for Risk and Innovation (GCRI Canada) is a Canadian not-for-profit organisation, not a registered charity, with its registered office in Ontario.