Executive Summary

Front matter 1 of 8 · ≈ 9 min read · 28 paragraphs · section 1 of 40 · search the book

This summary is intended for public leaders, heads of international institutions, central banks, donors, parliamentarians, businesses and research partners. It distinguishes the decisions that the present work makes it possible to take now from those that still require validation.¶ 1

1 The strategic problem

Decisions have changed scale. A monetary policy adopted in a large jurisdiction alters, elsewhere, the cost of financing, capital flows, the exchange rate, central bank responses, fiscal space and investment. A digital infrastructure may exert effects on water, electricity, data, labour, competition and access to public services beyond the territory that authorises it. A value chain may concentrate value in certain jurisdictions and displace social or environmental costs towards others. Benefits, risks and response capacities no longer coincide in space, in time, or across categories of actors.¶ 2

The problem does not lie first of all in intention. Upstream institutions act according to domestic mandates, defined objectives, models and accountability obligations. Downstream actors observe the effects that materialise. Both perspectives may be legitimate and incomplete. Credible governance must therefore replace the imputation of intent with a common procedure: documenting what was decided, what was transmitted, through which mechanisms, with what uncertainties, who benefited, who bore the cost, which internal vulnerabilities amplified the shock and what consequence is admissible in the light of the level of proof and of the applicable law.¶ 3

2 The diagnosis

The international system is dense but discontinuous. It produces norms, data, financing, sectoral mechanisms and specialised forums. It rarely connects the chain as a whole while preserving the evidence, the uncertainty, the objections and the correction pathways. The internal coding of 99 primary instruments and of 130 Sevilla initiatives is consistent with this hypothesis: financing and standard-setting are strongly represented; cross-cutting qualification, continuity towards a consequence and the organisation of redress much less so. This finding remains provisional pending double coding and the full opening of the corpora.¶ 4

The gap is therefore not simply “more institutions are needed”. It is functional and procedural: who measures? who qualifies? who decides on a consequence? who finances? who certifies the systems that produce the evidence? who answers before which forum? what becomes of the uncertainty when a case file passes from one actor to another?¶ 5

3 The central proposal

The Global Accountability Chain provides a grammar for answering these questions. It rests on three completenesses. Functional completeness verifies that the necessary transformations are covered. Relational completeness identifies, for each duty-holder, the forum before which it must explain and justify its conduct. Procedural completeness protects the case file as it crosses the interfaces: source data, assumptions, versions, uncertainties, dissents and redress.¶ 6

This grammar is reinforced by four safeguards: negative mandates, which prohibit the combination of the most conflict-prone roles; stopping rules, which permit not attributing or not pricing; comparison with lighter architectures; and the strict distinction between fact, causality, equity and law.¶ 7

4 The 7+3: what it is and what it is not

The 7+3 is a first candidate instantiation. It distinguishes seven functions: to measure; to qualify; to determine the admissible consequence; to apply certain adjustments; to finance prevention, transition or solidarity; to certify digital systems; to organise redress. Three instruments frame them: the Carthage Charter, the Global Accountability Framework and the Call of Carthage.¶ 8

The 7+3 is not the theory itself, does not impose seven new organisations and does not presume that every case file passes through every link. A function may be hosted by an existing institution, a network, a panel, a protocol or a distinct body. An architecture with three or five bodies, or a federated protocol, must be preferred if it protects evidence, independence and redress equally well at lower cost.¶ 9

5 A two-loop architecture

The Global Accountability Chain deals with effects that are observed or sufficiently measurable. The Preventive Accountability Loop acts upstream: anticipating a cascade risk, detecting exposures and thresholds, qualifying the alert, notifying decision-makers, obtaining a reasoned response, preparing a possible anticipatory action and evaluating the results. This loop is not an eighth institution. It distributes responsibilities: HuIB or an accredited network produces the prospective evidence; GIRA qualifies the alert level; the GAF organises notification and response; the GSF may support an anticipatory action; the competent forum reviews the procedure.¶ 10

The duty of prevention is procedural. It does not oblige a central bank, a government or a company to follow an external recommendation; it requires that reasonably foreseeable risks be able to enter into the decision, that the response be reasoned and that the quality of the alerts be evaluated. The loop must be rejected if it leads to government by forecast, to unstable alerts, to self-fulfilling panic or to a burden greater than its value.¶ 11

6 The monetary programme

The monetary programme is one of the most distinctive contributions, but it must be presented as an offer of empirical collaboration, not as an established causal result. It decomposes spillovers into fifteen signed mechanisms, includes offsets, distinguishes flows from stocks, and prevents double counting through allocation to primitive causal components. It analyses the superposition of instruments, the synchronisation of source jurisdictions, the defensive reactions of recipient countries and the exhaustion of fiscal and financial buffers.¶ 12

The shock is conceived as a multidimensional vector with testable signatures. Attribution is partial, bounded and conditional on a coherent counterfactual. The gains from easing and the costs of tightening are compared over the full cycle, by agent and by mechanism. The “100 units injected” protocol does not naively track a quantity of money; it distinguishes reserve creation, balance sheets, financing, asset valuation, incomes and exposures, which are never added together as a single budget.¶ 13

The Monetary Cascade Early-Warning Protocol extends this programme: it can inform decision-makers before the stacking of interest rates, quantitative tightening, exchange-rate reactions and the exhaustion of buffers produces a systemic contraction. Its function is to make the risk visible and to prepare the responses, not to set a global interest rate.¶ 14

7 The other applications

The artificial intelligence data centre constitutes an integrated demonstrator: it tests whether a case file can cross the functions without losing provenance, uncertainty or redress. It does not prove effectiveness in the field.¶ 15

The application to Sevilla proposes an actionability layer for development financing: initiative passport, maturity scale, decision gates, financial dictionary, tracking of useful flows, corrective loop and portfolio pilot. It can create operational value without waiting for the whole of the 7+3.¶ 16

Preventive peace is an exploratory extension. It studies how an undocumented grievance, with no forum and no correction mechanism, can contribute to mistrust and escalation. It does not claim that more measurement resolves conflicts grounded in incompatible interests or in power relations.¶ 17

8 Artificial intelligence, data and crisis

Artificial intelligence can reduce measurement costs, detect inconsistencies and bring analytical capacities closer together. It can also concentrate data, suppliers and the power to define the categories. The framework therefore distinguishes assistance, recommendation, validation and decision; no official alert, no pricing, no financing and no jurisdictional decision may be triggered automatically.¶ 18

Resilience rests on pre-commitment: thresholds, fast-track procedures, pre-arranged financing, expiry and oversight. Speed does not authorise the disappearance of provenance, of the human signature, of contestation or of operation without AI.¶ 19

9 What can be asserted now

The present work can maintain that it proposes an explicit theory of the three completenesses, a doctrine of interfaces, negative mandates, stopping rules, a function-first method, a set of pre-specified models and protocols, as well as a candidate architecture submitted to comparison.¶ 20

It can maintain that the available internal coding is consistent with a functional asymmetry and that the monetary programme provides a more disciplined decomposition than unaudited aggregate estimates.¶ 21

It can maintain that prevention, AI and resilience are integrated in the form of procedural duties, authority matrices and rejection criteria.¶ 22

10 What must not yet be asserted

The present work does not demonstrate that the seven functions are exhaustive, that seven organisations are necessary, that the 7+3 is superior, that the monetary amounts are causally validated, that a general legal obligation to prevent spillovers exists, that the instruments are compatible with all applicable bodies of law, that the preventive loop improves decisions, that the AI prototype is safe or that the architecture actually reduces conflicts.¶ 23

Zero-residual reconciliation is a form of accounting consistency; the corpora remain coded principally in-house; the demonstrator is a simulation; the full cost and the political economy of adherence are not established.¶ 24

11 Recommended programme of action

The next phase must replace internal improvement of the text with organised external validation: publishing the corpora, codebooks, scripts and version logs; having the double coding carried out; executing and replicating the monetary estimation; obtaining legal opinions act by act; costing the minimal, federated and integrated configurations; conducting a three-arm pilot; auditing the AI systems; financing the independent counter-assessment of affected parties; and mandating a red team tasked with searching for the conditions of failure.¶ 25

For general academic communication, the project must be led by the most mature institutional core. For the IMF, central banks, the BIS and econometric teams, the monetary programme must be presented as a protocol to be executed jointly. For decision-makers, the entry point is functional discontinuity, modularity and the pilot. For development partners, the entry point is the actionability of Sevilla.¶ 26

Decision table — status of the main proposals

Approval pathway. The project does not call for the immediate adoption of the 7+3. It proposes a graduated trajectory: external validation and the opening of the evidence; technical incubation within the development financing process; a procedural General Assembly resolution authorising a report and comparative pilots; incorporation of the validated functions into participating institutions or legal systems; then, only where the lighter mechanisms are insufficient, negotiation of legally binding modules. This trajectory is designed to complement, and not to reproduce, the commissions of 2009 and of 2025: it bears on execution, interfaces and redress, not on a new general list of reforms (§12.7).¶ 27

Executive conclusion

The Humanity Compact does not ask that the whole world immediately adopt a new architecture. It asks first that every decision liable to displace significant effects be open to examination through a common grammar: what evidence, what causality, which beneficiaries, which vulnerabilities, what admissible consequence, what prevention, what financing and what redress? Its value will be demonstrated not by the scale of its institutions, but by its capacity to produce more reliable information, a better reasoned decision, better preparedness and more effective redress than the alternatives, at a proportionate cost. The first reasonable commitment is therefore not the adoption of the 7+3; it is the opening of the evidence, methodological cooperation and a comparative pilot designed to be able to conclude that the architecture must be maintained, simplified or abandoned.¶ 28

@incollection{yaiche_hc_executive_summary,
  author    = {Yaïche, Nizar},
  title     = {Executive Summary},
  booktitle = {The Humanity Compact},
  year      = {2026},
  version   = {2.3},
  doi       = {10.5281/zenodo.22746533},
  publisher = {Zenodo},
  url       = {https://humanitycompact.org/ouvrage/executive-summary/}
}

A book by Nizar Yaïche, former Minister of Finance of the Republic of Tunisia. Who is the author?