How AIIM Can Help Implement UNESCO Open Science Principles

Getting your Trinity Audio player ready...

The AI Internet-Meritocracy (AIIM) can help governments implement UNESCO open science principles by connecting public funding to documented open contributions. Instead of merely asking researchers to publish openly, AIIM can recognize and reward open articles, datasets, software, methods, replications, reviews, and educational resources.

This converts open science from a declaration into an operational incentive system.

UNESCO’s Recommendation on Open Science calls for accessible knowledge, transparent scientific processes, equitable participation, sustainable infrastructure, collaboration, and reform of scientific incentives. AIIM cannot implement every part of that programme by itself. It can, however, address one of its most persistent obstacles:

Researchers are often told to practise open science while their careers and funding remain governed by credentials, institutional prestige, journal rankings, and closed committee decisions.

AIIM proposes a different mechanism. Researchers and developers could be evaluated from documented public contributions rather than primarily from degrees, affiliations, grant-writing ability, or access to influential institutions.

This article develops the argument made in the previous article, [How Every Government in the World Failed to Follow UNESCO Recommendations for Open Science — insert published URL]. That article examined the gap between governments’ formal support for open science and the continued operation of scientific systems that remain institutionally closed. The present article explains how AIIM could help close that implementation gap.

What UNESCO Means by Open Science

The UNESCO Recommendation on Open Science was adopted on November 23, 2021. UNESCO describes open science as a set of principles and practices intended to make scientific knowledge accessible to everyone while making the production and evaluation of knowledge more inclusive, equitable, transparent, and sustainable.

UNESCO identifies four central values:

• quality and integrity;
• collective benefit;
• equity and fairness;
• diversity and inclusiveness.

It also identifies guiding principles including transparency, scrutiny, reproducibility, equality of opportunity, accountability, collaboration, flexibility, and sustainability.

Open science therefore means much more than removing a paywall from a journal article. It includes:

• open scientific publications;
• open research data and metadata;
• open-source software and source code;
• open hardware;
• open educational resources;
• public engagement and citizen science;
• transparent research evaluation;
• equitable access to participation and scientific benefits.

UNESCO asks Member States to act in seven broad areas: policy, infrastructure, capacity-building, incentives, innovative methods, shared understanding, and international cooperation. Its implementation programme also includes work specifically devoted to funding and incentives, infrastructure, monitoring, policies, and capacity-building.

Why Open Science Policies Often Remain Symbolic

A government can adopt an open-access policy without changing the economic structure of science.

For example, it may require publicly funded articles to be deposited in repositories while continuing to:

• allocate grants through closed committees;
• use university rank as a proxy for research quality;
• reward publication in prestigious journals more than reusable research outputs;
• exclude researchers without conventional institutional positions;
• treat data, software, replication, reviewing, and maintenance as secondary work;
• concentrate funding in already successful institutions;
• measure compliance without rewarding the people who perform the additional work.

This creates an incentive contradiction.

Researchers may be asked to document data carefully, maintain software, publish negative results, prepare reusable metadata, answer public criticism, and preserve outputs for many years. Yet career systems may continue to reward only papers, citations, degrees, and institutional advancement.

Open science cannot become sustainable when openness creates more work but little additional recognition or funding.

A durable open science policy must reward the production, maintenance, explanation, verification, and reuse of open knowledge—not merely require its deposit.

What AIIM Adds to Open Science

AI Internet-Meritocracy is an experimental system designed to evaluate documented scientific and open-source contributions and distribute funding accordingly.

Its central policy idea is simple:

Publicly useful research contributions should create a credible claim to financial support, even when the contributor lacks a prestigious affiliation, a conventional career position, or a successful grant proposal.

AIIM could examine connected public records such as ORCID profiles, publications, repositories, datasets, software projects, documented dependencies, reviews, and other research outputs. It could then generate an assessment that is subject to disclosure, correction, dispute, and audit.

The present AIIM beta remains experimental. Its evaluations are heuristic rather than validated measurements of scientific or economic value, and some planned decentralization and on-chain functions are still under development. This distinction is important for any government considering a pilot programme.

AIIM should therefore be regarded as proposed open-science infrastructure—not as an infallible automated judge.

Mapping UNESCO Implementation Priorities to AIIM

UNESCO implementation priority	Possible AIIM contribution
Shared understanding of open science	Publish explicit definitions of eligible open outputs and explain how each output affects evaluation
Enabling policy environment	Provide a configurable funding layer that governments can connect to national open-science policies
Infrastructure and services	Connect persistent identities, repositories, publications, datasets, source code, licences, and payment records
Training and capacity-building	Reward documentation, educational resources, mentoring, translation, metadata improvement, and community support
Culture and incentives	Pay for useful open contributions rather than treating openness as an unfunded obligation
Innovative approaches	Use AI-assisted evaluation, contribution graphs, continuous funding, public evidence, and auditable decision records
International cooperation	Support global and national funding pools using compatible evaluation infrastructure

This mapping does not imply that AIIM alone satisfies the Recommendation. It shows where AIIM could become an implementation component within a broader public open-science programme.

Rewarding Open Outputs, Not Only Journal Articles

One of AIIM’s most important potential contributions is widening the category of work that receives recognition.

Modern research depends on much more than finished articles. It also depends on:

• datasets;
• research software;
• maintained libraries;
• laboratory protocols;
• machine-readable metadata;
• formal proofs;
• replications;
• negative results;
• peer reviews;
• translations;
• educational explanations;
• long-term archives;
• corrections and retractions;
• community infrastructure.

The FAIR Guiding Principles emphasize that research objects should be Findable, Accessible, Interoperable, and Reusable. The original FAIR publication also explains that these principles can apply to algorithms, tools, and workflows, not only conventional datasets.

AIIM could create an economic incentive for FAIR implementation by recognizing work that improves:

• persistent identification;
• metadata completeness;
• provenance;
• licensing clarity;
• interoperability;
• machine readability;
• documentation;
• long-term reusability.

A dataset that nobody can find or interpret is technically published but functionally closed. A software repository without documentation, dependencies, versioning, or maintenance may be nominally open source but difficult to reuse.

AIIM could assess degrees of practical openness rather than relying on a binary open-or-closed label.

Making Scientific Evaluation More Transparent

UNESCO associates open science with scrutiny, critique, reproducibility, responsibility, and accountability. These principles apply not only to research outputs but also to the institutions that evaluate them.

Traditional grant systems often disclose the names of successful projects but not the complete reasoning, evidence, uncertainty, or alternatives considered during evaluation.

An open AIIM implementation could publish:

• the evidence considered;
• the criteria applied;
• the version of the evaluation model;
• uncertainty or confidence information;
• detected conflicts and missing information;
• explanations for allocations;
• changes resulting from appeals;
• public audit records;
• aggregate bias and error analyses.

This would not automatically make the evaluation correct. It would make it more inspectable.

The distinction matters:

Transparency does not eliminate error, but it allows error to be identified, disputed, measured, and corrected.

Government deployments should therefore require explainable assessments, versioned criteria, reproducible evaluation procedures, and a meaningful right of appeal.

Human review should remain available for ambiguous, high-stakes, disputed, or safety-sensitive cases. Human governance is also needed to respond to prompt injection, fabricated evidence, coordinated manipulation, identity disputes, and attempts to game evaluation criteria.

The proposed adversarial testing of AIIM is relevant because an open-science funding system must itself be open to scrutiny.

Aligning Research Assessment With Open Science

Open science cannot succeed while assessment systems reward researchers mainly for journal prestige.

The San Francisco Declaration on Research Assessment argues that research outputs should be evaluated on their own merits rather than through journal-level metrics. The Agreement on Reforming Research Assessment similarly calls for recognition of diverse research outputs, practices, and activities, using qualitative judgement supported by responsible quantitative indicators.

AIIM could support this reform by evaluating a wider evidence base:

• what the contributor produced;
• whether it is publicly inspectable;
• whether others can reuse it;
• what later work depends on it;
• whether the contributor maintains or corrects it;
• whether the work improves scientific infrastructure;
• whether independent communities find it useful;
• what uncertainty remains about its value.

Citation counts may be relevant evidence, but they should not become the sole definition of merit. Citation patterns vary drastically among disciplines and may reproduce existing inequalities.

A credible system must distinguish visibility from value.

AIIM and the Three-Degree System

The bachelor’s–master’s–doctorate system does not inherently contradict UNESCO open science principles. Degrees can certify that a person completed a course of study or demonstrated certain competencies.

The conflict begins when a degree becomes a universal gate controlling who may:

• receive research funding;
• submit work for serious evaluation;
• lead a project;
• obtain access to infrastructure;
• be treated as a scientist;
• receive recognition for a valid contribution.

UNESCO’s principle of equality of opportunity is difficult to reconcile with a system that automatically excludes contributors because they lack a doctorate, institutional position, or conventional academic career.

AIIM offers a narrower and more practical correction. It does not need to abolish universities or degrees. It can treat qualifications as one form of evidence without treating them as an absolute prerequisite.

Under this model:

• a degree may support an assessment;
• institutional experience may remain relevant;
• demonstrated expertise may matter;
• but published and verifiable contributions remain directly assessable;
• lack of a degree does not automatically erase the contribution.

This is particularly important for independent researchers, citizen scientists, free-software developers, data stewards, retired scientists, researchers displaced by war or migration, and contributors from regions with limited access to doctoral education.

The AIIM model for non-PhD contributors is therefore not merely a social inclusion feature. It is a mechanism for implementing UNESCO’s equality-of-opportunity principle.

Persistent Identities and Contribution Records

Open science requires reliable links between people and their contributions.

ORCID provides persistent researcher identifiers and records that can connect researchers with activities and outputs across institutions, disciplines, and borders.

An AIIM implementation could use ORCID and compatible systems to reduce dependence on institutional affiliation. It could also connect:

• publication identifiers;
• dataset identifiers;
• software repositories;
• contributor roles;
• funding records;
• review activity;
• corrections;
• declared conflicts of interest;
• dependency relationships.

Governments should avoid creating another isolated national profile database where international infrastructure already exists. AIIM should interoperate with open standards and allow researchers to correct, export, and control their records.

A person’s scientific identity should not disappear when that person changes university, loses employment, migrates, or works independently.

Funding Open Science Continuously

Traditional grants normally provide money before the final contribution exists. Evaluation must therefore rely heavily on promises, plans, credentials, and institutional trust.

AIIM could add a complementary mechanism: retrospective and continuous funding based on demonstrated public contributions.

This could support work that conventional grants handle poorly:

• maintaining mature scientific software;
• updating datasets;
• correcting old publications;
• answering reuse questions;
• improving documentation;
• preserving digital resources;
• translating scientific materials;
• replicating results;
• reviewing overlooked research;
• integrating outputs created by different groups.

Continuous support is especially relevant to scientific infrastructure. A repository, database, library, or standard may remain useful for decades, but its value can be destroyed when maintenance funding ends.

AIIM could periodically reassess documented use, dependencies, maintenance, and public benefit rather than treating funding as a single competition with a fixed endpoint.

Combining National and Global Funding

UNESCO describes science as serving collective benefit and seeks to reduce technological and knowledge gaps between and within countries.

However, governments remain accountable to their own populations and may legitimately fund national priorities.

AIIM could support both objectives through separate but interoperable funding pools:

• a global open-science pool;
• national research pools;
• regional pools;
• disciplinary pools;
• public-health or climate pools;
• charitable and philanthropic pools;
• infrastructure-maintenance pools.

The same open contribution could be considered by several pools under different mandates. A national government might support domestic researchers, while an international fund rewards the global usefulness of the same work.

This federated approach is described further in National Science Policy vs Science as a Global Public Good.

Common infrastructure would make contributions discoverable and assessable across borders, while each funder retained lawful control over its eligibility rules and policy priorities.

Monitoring Real Implementation

UNESCO Member States agreed to report periodically on their implementation of the Recommendation. UNESCO has also developed implementation working groups and a global monitoring framework.

AIIM could contribute evidence for such reporting by producing aggregate indicators such as:

• the proportion of funded outputs that are openly accessible;
• the proportion carrying persistent identifiers;
• availability of data, code, methods, and licences;
• funding received by independent researchers;
• geographic and linguistic distribution of recipients;
• support for research software and infrastructure;
• money allocated to replication and verification;
• time between publication and recognition;
• successful appeals and corrected evaluations;
• concentration of funding by institution or career stage.

These indicators should not become simplistic targets. A government could increase its number of open files while failing to improve their quality, accessibility, or reuse.

Monitoring should therefore combine quantitative evidence with qualitative auditing.

The objective is not to maximize the number of objects labelled “open.” It is to increase the amount of scientific knowledge that people and machines can legitimately discover, inspect, understand, verify, and reuse.

Applications Beyond Government

Although governments control large research budgets, UNESCO open science principles apply to a wider ecosystem.

Universities

Universities could use AIIM reports as one input when recognizing datasets, software, reviewing, replication, public engagement, and infrastructure work.

They should not use an AIIM score as an automatic hiring or dismissal mechanism. Its appropriate role is to broaden the evidence considered, not to replace contextual academic judgement.

Research funders

Public agencies and foundations could reserve part of their budgets for continuous retrospective funding of demonstrated open contributions.

This would complement proposal-based grants rather than immediately replacing every existing funding method.

Charities and donors

Charities could create thematic pools for open medical research, mathematical infrastructure, climate data, neglected diseases, scientific software, or research in underfunded regions.

Journals and repositories

Repositories could provide structured metadata about licences, versions, provenance, linked code, datasets, reviews, and corrections. Journals could expose transparent contribution data rather than limiting recognition to authorship order.

Scientific communities

Professional societies and research communities could help define discipline-sensitive evaluation criteria. What counts as reuse or impact in pure mathematics differs from what counts in clinical research, archaeology, software engineering, or linguistics.

International organizations

International organizations could operate cross-border funds and use common AIIM infrastructure without requiring a single centralized world authority.

A Government Implementation Model

A government does not need to transfer its entire research budget to an experimental system.

A responsible implementation could begin with a limited pilot.

Establish a legally defined open-science fund

The fund should specify eligible outputs, public-interest objectives, privacy requirements, exclusions, and appeal rights.

Make open outputs machine-discoverable

Require persistent identifiers, structured metadata, contribution statements, licences, provenance information, and links between articles, data, software, and funding.

Connect public contribution records

Use interoperable infrastructure such as ORCID, DOI registries, repositories, and software forges rather than building a closed national identity silo.

Run AI-assisted evaluations

Assess documented contributions while recording the model version, evidence, explanation, uncertainty, and detected limitations.

Add human oversight and appeals

Create procedures for missing outputs, mistaken identity, disciplinary misunderstanding, model bias, security concerns, and disputed authorship.

Publish allocation and monitoring data

Release auditable records while protecting personal data, confidential research, endangered species information, security-sensitive knowledge, and legitimate Indigenous knowledge restrictions.

UNESCO’s principle is “as open as possible,” not indiscriminate disclosure of everything.

Commission independent audits

Test the system for regional, gender, linguistic, institutional, disciplinary, and career-stage bias. Conduct adversarial testing for fabricated records, citation manipulation, prompt injection, and coordinated voting attacks.

Expand only after evidence

A pilot should be expanded only when it demonstrates acceptable accuracy, procedural fairness, security, administrative cost, and public value.

The existing AIIM proposal for governments describes possible national portals and integration with existing ministries. Any public deployment should, however, preserve the experimental safeguards and limitations stated above.

What AIIM Cannot Do Alone

AIIM is primarily a funding, evaluation, and accountability mechanism. It cannot by itself:

• build every repository and broadband network;
• guarantee the correctness of research;
• replace scientific expertise;
• determine the appropriate openness of sensitive data;
• reform copyright and data-protection law;
• preserve every research object indefinitely;
• eliminate bias from incomplete public records;
• prevent every form of gaming;
• solve inequality in scientific education;
• create international trust automatically.

AIIM could even reproduce existing inequality if its input data disproportionately represent English-language, well-indexed, highly connected, or institutionally privileged researchers.

That risk must be treated as a core design problem.

Researchers should be able to add missing evidence, contest factual errors, request reconsideration, and understand the main reasons behind an assessment. Governments should periodically test whether the system mistakes visibility, prestige, or digital accessibility for scientific merit.

From Recommendation to Infrastructure

The UNESCO Recommendation established an important international consensus. But principles do not implement themselves.

Governments and institutions need systems that make openness:

• professionally valuable;
• financially supported;
• technically measurable;
• publicly accountable;
• accessible beyond established institutions;
• compatible with global scientific cooperation.

AIIM could become one part of that infrastructure.

Its most important contribution is not merely the use of artificial intelligence or blockchain. Its deeper contribution is the proposed connection between open contribution and material recognition.

Open science becomes sustainable when people who create, maintain, explain, verify, and share public knowledge are directly recognized for doing so.

For governments, AIIM offers a possible bridge between an international recommendation and an operating funding mechanism. For universities, charities, journals, repositories, scientific communities, and international organizations, it offers a common framework for rewarding contributions that the current system frequently treats as invisible.

AIIM should not be adopted uncritically. It should be piloted, audited, challenged, corrected, and governed transparently.

That process would itself embody the UNESCO principles AIIM is intended to advance.