INDEPENDENT POLICY & RESEARCHConceptual framework Edition 0.16.4
Present-day tabletop scenario

Three-country oil-supply warning

Several reports repeat one unconfirmed oil claim. Independent source checking could give all three countries a more reliable basis for discussion.

Hypothetical exercise · Reviewed 9 November 2026

Reviewers identify source dependencies and preserve uncertainty. The exercise asks whether a correction changes the quality of a civilian briefing; it does not model market prices or a government’s future response.

Questions worth examining

Generated tabletop brief · Mutual AI Resilience

United States · China · Russia: crude oil & refinery planning

Disputed supply warning · 3-day exercise · Maintain essential civilian continuity

MARHUMAN DECISIONSU.S.Authority stays localChinaAuthority stays localRussiaAuthority stays local
Illustrative information relationship · no live systems connected

The scenario

A widely repeated civilian supply claim appears to come from one original source. Independent confirmation is missing.

The question: Which observations are independent, and which reports repeat the same claim?

Crude is an input to refining. A crude quantity is not a finished-fuel delivery; processing, suitability, permissions, and physical capacity remain separate questions.

What each country could gain

These are conditional gains to test, not observed national results. Participation does not require a commercial transaction between every country.

United States

The possible gain

Civilian analysts and permitted buyers could reconcile conflicting public supply reports and distinguish a market-wide concern from a verified change relevant to their own planning.

Contribution, authority, and trade-off
What it could contribute

Approved civilian demand summaries, product requirements, public disruption notices, and human-reviewed corrections.

What it would retain

U.S. authorities and operators retain regulatory, commercial, safety, and participation decisions.

What needs scrutiny

Protect commercial information, avoid reliance on an unverified foreign report, and account for the staff needed to review AI output.

China

The possible gain

Permitted buyers could check whether an approved supply description matches their stated requirements and compare revisions before committing to an assumption.

Contribution, authority, and trade-off
What it could contribute

Approved buyer or supplier requirements, civilian logistics summaries, and clarification of permitted product and document categories.

What it would retain

Chinese authorities and operators retain domestic supply, import, export, safety, and information-sharing decisions.

What needs scrutiny

Protect commercial information and domestic priorities; assess whether the process shifts delay or cost onto Chinese participants.

Russia

The possible gain

Permitted sellers and civilian planners could obtain a clearer record of accepted requirements and unresolved questions, reducing avoidable documentary disputes without assuming finance or export permission.

Contribution, authority, and trade-off
What it could contribute

Where permitted, approved civilian needs or offers, document clarifications, and confirmation of what has actually been accepted or received.

What it would retain

Russian authorities and operators retain their decisions; every other participant retains its own applicable legal and commercial requirements.

What needs scrutiny

Eligibility, counterparties, services, finance, and information access need separate review. An offer must not be treated as permission or delivered supply.

How cooperation could reduce escalation

Repeated claims can look like corroboration and prompt public blame. Checking provenance could create room for clarification before consequential political choices.

The proposed intermediate result is better-supported clarification and a usable civilian channel. A reduction in the chance of war would remain unproven.

Examine the assumptions

Evidence available
One unconfirmed source
Information-sharing boundary
Public information only
Working relationship
Limited working contact
Legal & institutional authority
Not established for this exercise
Local independence
Essential local functions retain a fallback
Primary civilian objective
Maintain essential civilian continuity

Review. Authority is not established. A hypothetical brief does not authorize participation, disclosure, payment, transport, or a transaction.

Review. One source is unconfirmed. Repetition does not establish corroboration; preserve the uncertainty before a consequential decision.

What participants could share

  • Only public records with source, observation time, revisions, and uncertainty.
  • A public claim does not establish an accepted commercial commitment.

Share the minimum approved information. A public source is not a substitute for authority to use or redistribute information for a new purpose.

The cooperation process

The same authorized record and human decisions, with AI-assisted document comparison, reviewed translation, and inconsistency checks.

  1. Agree on the civilian question, participants, permitted records, and human decision rights.
  2. Preserve source dependencies and present unconfirmed claims as unconfirmed.
  3. Use AI to draft comparisons or reviewed translations; check every material claim against original records.
  4. Allow each participant to challenge a finding and record unresolved disagreement.
  5. Authorize local decisions separately, confirm actual receipt or continuity, and test the fallback.

The role of AI

AI drafts or flags material for responsible human review; originals and competing accounts remain available.

The trade-off

Potential review assistance comes with model error, automation bias, provider dependence, and added verification costs.

Optional quantity check

Optional quantity check: enter demand, confirmed usable supply, and usable delivery capacity in the same daily unit. The tool will show transparent arithmetic from your inputs.

What would show that it helped?

  • Confirmed usable input, unresolved specification questions, avoidable administrative delay, and the separate path to finished products.
  • Unsupported attributions, duplicated sources counted as evidence, and time to correction.
  • Unsupported accusations, correction time, and whether human recipients acknowledge the correction.
  • Review cost, disclosure errors, unequal burdens, and local performance when the shared channel is removed.

Compare current exercise practice, structured cooperation without AI, and MAR with the same evidence, physical resources, information-release times, and evaluation criteria.

What this brief cannot establish

  • These gains and escalation pathways are proposed, not observed national outcomes.
  • A clearer record cannot create supply, transport capacity, permission, political consent, or product suitability.
  • No automatic price, revenue, recovery-time, or war-risk improvement is assigned to AI.
  • Compare the same evidence and physical resources across all three approaches.