Part I — Situation overview

On Thursday, 17 September 2026, the Prime Minister’s Office filed a criminal complaint against persons unknown with the National Bureau of Investigation over the state funding of the Artificial Intelligence Research Centre Nonprofit Ltd. (Mesterséges Intelligencia Kutató Központ, MIKK). According to the ministry’s statement, there is suspicion of budget fraud, an agreement restricting competition in a public procurement procedure, breach of trust and abuse of office. The statement was reported consistently by Telex, 24.hu, HVG, 444.hu and Népszava. According to it, the company, founded in December 2024, received 300 million forints barely two months later, on 12 February 2025, through Bethlen Gábor Fund Management (Bethlen Gábor Alapkezelő Zrt.), on the proposal of Vince Szalay-Bobrovniczky, then deputy state secretary for social affairs at the Prime Minister’s Office. The purpose of the funding was research into the integration of artificial intelligence into the state and society. According to the Prime Minister’s Office, the company had neither an operating history nor proven research capacity. The ministry also claims that the substantive research was contracted out through public procurement to a media-monitoring company for a gross 152.4 million forints, and that some of the submitted material appears to be machine-generated. The company transferred back a further 300 million awarded without competition for 2026 on 14 April, two days after the election. Vince Szalay-Bobrovniczky called the complaint, which also extends to breach of trust, a “witch hunt”. According to him, before making the proposal he personally consulted the rector of the university employing the lead researcher, and the reason for the repayment was not a lack of capacity but that the researchers saw no guarantee that their work could continue. A criminal complaint is not a suspicion: the investigating authority decides whether to open an investigation, and the former deputy state secretary is not a suspect.

On the same day Zoltán Tanács, Minister for Science and Technology, outlined a new direction for research policy on InfoRádió and at a background briefing. According to Portfolio’s summary, Hungary currently spends around 1.3–1.4 per cent of gross domestic product (GDP) on research and development (R&D); this would be raised above 2 per cent within four years, and to around 3 per cent over eight years. The HUN-REN Hungarian Research Network would not be restored in its pre-2019 form, but with hybrid governance: researcher self-governance, the Hungarian Academy of Sciences and the government would all be given a role. The network would essentially be attached to the Academy, and the social science institutes transferred to ELTE would return to it. According to Telex, the minister asked András Stipsicz, director general of the Alfréd Rényi Institute of Mathematics, to be interim president, but the decision falls within the competence of HUN-REN’s Governing Board and Supervisory Board. The minister hopes the transition will last two to three months, until the adoption of the bill due to be ready in the autumn. MIAK has already proposed competition-based research funding built on independent professional review when the Krausz Élvonal contract was terminated. And in early May it urged that academic autonomy and HUN-REN be settled in a cardinal act.

MIAK’s reading: the two news items are two sides of the same question. Whether a crime was committed will ultimately be decided by a court after the investigation, but the structure of the decision can be judged from both the Prime Minister’s Office’s statement and the former deputy state secretary’s response. A fund not specialised in research funding decided on public money for research, on the proposal of a single official, without a public call for proposals or independent review. Even the personal consultation described as made in good faith shows precisely that control depended on one person, not on an institution. If R&D spending rises by a further 0.7 percentage points of GDP but this channel stays open, more money also means more risk. The credibility of research funding comes not from the amount, but from the rule of allocation.

Part II — Foundations in the literature

Three works provide a framework for the three questions of the topic. Robert Klitgaard (American economist, author of the structural model of corruption) writes in his book Controlling Corruption (1988) that abuse flourishes where the decision-maker has a monopoly and broad discretion, and accountability is weak. The MIKK funding met all three conditions. Klitgaard also shows that the US government screens applicants’ eligibility before awarding funding, and that semi-autonomous, hybrid organisations under public oversight can serve both performance and the public interest. According to the book 23 Things They Don’t Tell You About Capitalism (2010) by Ha-Joon Chang (South Korean-born economist, researcher of the developmental state and industrial policy), in every developed market economy the state funds a significant share, 20–50 per cent, of R&D. The 2 per cent target will therefore not be met without a state commitment, but it is the quality of the allocation of state money that determines its effect. Peter F. Drucker (Austrian–American management thinker, founder of the methodology of measuring organisational effectiveness) writes in The Effective Executive (1967) that feedback must be built into every decision, measuring expectations against actual events. He also warns that research priorities geared to quick success do not produce breakthroughs. The detailed treatment of the literature — by author, with quotations — is in section 6.4 Literature in detail.

Part III — MIAK’s concrete proposal

MIAK proposes three measurable measures. The first settles the channel for allocating public research money, the second the measurability of the 2 per cent target, and the third the guarantees of the restructuring of HUN-REN. All three can be built into the legislation now being prepared and into the medium-term budget plan.

3.1 Public research money only through a research-funding channel: open call, independent review, subcontracting cap, public results (legislative amendment within 90 days)

MIAK proposes that legislation state that state funding for research purposes above 20 million forints may only be awarded by a body established for research funding, primarily the National Research, Development and Innovation Office (NKFIH). The decision should be taken through an open call for proposals and independent professional review (peer review). Other funds and ministerial budget lines — including the Bethlen Gábor Fund — should not be able to give money for research by individual decision. If the government nevertheless exceptionally awards individual research funding, it should publish the justification and the opinions of at least two external reviewers before the decision. The minimum eligibility requirements for applicants: named researchers with a verifiable professional track record, and an operating history or a host research institution. Klitgaard (see 6.4.1) describes precisely this screening before funding is awarded. At MIKK, more than half of the funding (152.4 million of the 300 million) went to a subcontractor. MIAK therefore proposes that at most 30 per cent of research funding may be subcontracted without prior approval. The results of completed projects should be placed in an open-access public repository within six months. The beneficiary should also declare whether it used generative artificial intelligence tools, and for what. In this way anyone can check whether research took place. This is the research application of A2 public procurement transparency and the A1 public funds dashboard, while the declaration follows the transparency principle of the D1 responsible AI framework.

3.2 The annual path and output indicators of the 2 per cent target (together with the medium-term budget plan)

The minister’s four-year target — from 1.31 per cent in 2024 (Eurostat) to above 2 per cent — means an increase of roughly 0.7 percentage points of GDP. MIAK proposes that the government publish a year-by-year path in the medium-term budget plan. The path should show separately the expected contribution of the public sector (universities, research institutes) and of the business sector. According to Eurostat’s 2024 data, the gap is largest in public-sector research: university and government research spending together amount to 0.31 per cent of GDP, against an EU average of 0.72 per cent. The business sector achieves 0.99 per cent, against the EU’s 1.49 per cent. According to Chang’s data (see 6.4.2), in developed economies the state funds 20–50 per cent of R&D; the path must also lay down this state share in forints. Alongside the money, the plan should publish five output indicators with baseline values: the domestic share of the most internationally cited publications, the number of European patent applications, the number of companies spun off from research institutions (spin-offs), the ratio of private capital matching state funding, and the retention of young researchers. The private capital ratio directly measures the direction indicated by the minister, co-financing with private investors (G12). For funds earmarked for artificial intelligence research, productivity impact measurement under D12 should also apply. Because of Drucker’s warning (see 6.4.3), the indicators should be evaluated at portfolio level, over a five-year window, not project by project and annually. In this way they do not reward quick, safe projects. The path specifies the G4 innovation ecosystem programme point, and the annual evaluation specifies the ex-post impact assessment of G20.

3.3 Limits on HUN-REN’s transitional period and the statutory minimum of researcher self-governance (in the bill promised for the autumn)

The minister promises a two-to-three-month transition, but the powers of the transitional leadership are laid down nowhere. MIAK proposes that the Governing Board prescribe the limits when appointing the interim president. No irreversible structural decision should be taken during the transition: merging or closing research institutes, disposing of assets, or entering into long-term commitments. The appointment should have a fixed end date. The permanent president should be selected through an open, international call, on the basis of public selection criteria. The majority of the selection committee should consist of members elected by researchers and of foreign scientists. The law on the hybrid model should lay down three elements. The first is the composition of the board: members delegated by the researchers and the Academy should together be in the majority, and government members in the minority. The second is a statutory floor for multi-year funding, which, according to Portfolio’s report, the minister also intends to maintain. The third is a public international evaluation of the institutes, repeated every five years. Pay settlement is also urgent: the minister himself mentioned the pay gap between researchers who remained in the HUN-REN research network and those transferred to ELTE. And on 18 September the trade union of the Hungarian Research Centre for Linguistics wrote of an “unacceptable” pay situation. According to Klitgaard’s idea about hybrid organisations (see 6.4.1), the semi-autonomous construction works if public oversight is retained alongside autonomy. This is the science-policy application of the A6 checks and balances programme point.

The three proposals are linked by one principle: public money for research should not depend on the discretion of an individual official or minister, but on rules laid down in advance and made public. In Klitgaard’s formula this means narrowing discretion and expanding accountability; in Drucker’s language, feedback built into the decision. And Chang’s data remind us that the state should not withdraw from research funding, but should do it according to better rules.

Part IV — Expected effects and risks

Dimension Expected effect Risk
Transparency Public research money is channelled into a single call-for-proposals channel; in the public results repository anyone can check what was produced with the money The channel rule can be circumvented if the funding is called not “research” but an “expert study” or “communication”; the concept must be defined broadly
Science and research Independent review strengthens quality, public results strengthen citability; a limited transition increases researchers’ trust The review system is slower; in a small country it is hard to find reviewers without conflicts of interest, so foreign reviewers are needed, which is costly
Economy The annual path and the private capital ratio make the additional R&D spending predictable for companies too With narrow fiscal room for manoeuvre, the state share may suffer at the first round of cuts; the path does not protect against this, it only makes the shortfall visible
Public administration The limits on the interim leadership prevent irreversible decisions from being taken during the transition Too narrow transitional powers could paralyse day-to-day operations if the law is delayed

The main trade-off lies between speed and control. The call-for-proposals channel is slower than an individual decision. For urgent, genuinely strategic research, the government may therefore be tempted to make an exception. Proposal 3.1 does not prohibit this, but ties it to public justification and external review: an exception thus remains possible, but politically costly. The other risk is the distorting effect of indicators. If the funding of research institutions is tied annually to the number of publications, researchers will choose safe, short projects, and precisely the long-term, risky research for which the state provides funding in the first place will be lost. That is why MIAK proposes portfolio-level, five-year evaluation. In the case of HUN-REN, the proposal becomes risky if the law promised for the autumn is delayed and the transitional leadership operates with limited powers for months. It is therefore worth laying down, alongside the end date, a deadline for submitting the bill as well.

Part V — Measurability and summary

5.1 What is worth following? (suggested KPIs)

MIAK proposes four key performance indicators (KPIs), from which it will be visible after 12 and 24 months whether research funding has moved towards open competition and measurability:

  • Share of public research money awarded through calls for proposals: it is worth following what percentage of state funding for research purposes above 20 million forints was awarded through an open call with independent review. The proposed target from 2027 is 100 per cent, with the number of individual research decisions outside the research-funding channel at zero.
  • Share of public results: for what percentage of completed, state-funded research projects the results are publicly available within six months of completion; the proposed target is at least 90 per cent.
  • Keeping R&D spending on its path: whether GDP-proportionate spending according to Eurostat (2024: 1.31 per cent) reaches the value laid down in the medium-term plan each year. It is worth following the share of public-sector research separately (2024: 0.31 per cent), because that is where the gap is largest.
  • Closing HUN-REN’s transition: whether the transitional period ends within the announced two to three months, whether the permanent president is selected through an open call, and whether members elected by researchers and academic members are in the majority on the governing board established by the law.

5.2 Summary

MIAK asks the government and the National Assembly to close, before raising R&D spending, the channel through which the MIKK funding also passed. Public research money should only be awarded by a research-funding body, through an open call, with independent review, a subcontracting cap and public results. The 2 per cent target should be accompanied by an annual path and five output indicators with baseline values. And in the restructuring of HUN-REN, decisions should be taken not only about persons, but also about rules: a limited transition, an open call for the presidency, and a researcher majority laid down in law. The criminal complaint brings a case from the past before the authorities; the future will be decided by the rules of funding.

Two of MIAK’s foundational values are affected. Transparency, because public research money is credible if both the reason for the decision and the result are public. This is precisely what was missing in the MIKK case: the public could only learn of the accounts after the change of government, in the form of a criminal complaint. Data-drivenness, because the 2 per cent target in itself measures spending, not results. Only output indicators laid down with baseline values show whether more money has also brought more knowledge.

Part VI — Reasoning and further sources

6.1 Press framing by spectrum

In the left-liberal segment, 444.hu brought the prehistory of MIKK to the fore: the public knew the company from its having called fake an aerial photograph of one of the opposition’s events last year. The paper also highlighted that the office cannot be found at its registered address, and that the funding was decided in two days. It also covered the connections of the subcontractors: Observer is linked to Nézőpont, and the company renting out the property is an interest of a former Fidesz local councillor. Telex named the National Bureau of Investigation and the ministry led by Bálint Ruff, and recalled its own May article, according to which the second grant was given by Gergely Gulyás, then minister. At the same time, the paper’s headline names “the government” as the complainant, whereas in public-law terms the complaint was filed by the Prime Minister’s Office as a ministry. HVG published two separate articles on the complaint and on Vince Szalay-Bobrovniczky’s reaction. In the first it highlighted that the former deputy state secretary is the brother of the then defence minister; in the second it put the word “witch hunt” in the headline. Népszava gave a factual summary based on MTI.

In the public-affairs segment, 24.hu published the most complete documentation. It set out the details of the public procurement and published the former deputy state secretary’s response verbatim and in full. ATV framed the dispute as a “war” and gave more room to the former deputy state secretary’s arguments. In the economic segment, Portfolio did not cover the MIKK case. It did, however, discuss Zoltán Tanács’s programme in detail: hybrid governance, the 2 and 3 per cent R&D targets, the future of the Élvonal programme, the scaling back of non-repayable start-up grants and the decentralisation of state IT procurement.

In the conservative segment, Magyar Nemzet reported the change of leadership at HUN-REN citing Telex’s information, with the cautious wording “may have asked”. It highlighted the board’s competence over the decision, and that the ministry would replace the network’s entire leadership. The conservative segment did not cover the MIKK case among the articles examined. Common across the whole spectrum is that no one linked the two news items. Nor did anyone write about why a fund not specialised in research funding decided on research money. This channel question is at the centre of MIAK’s proposal.

6.2 Facts and data

Data Value Source
Criminal complaint 17 September 2026, against persons unknown, with the National Bureau of Investigation; complainant: Prime Minister’s Office Telex, 24.hu, Népszava, 17 September 2026
Founding of MIKK December 2024 Telex, 24.hu, Népszava
First grant HUF 300 million, 12 February 2025, through Bethlen Gábor Fund Management, on the proposal of the deputy state secretary for social affairs Telex, 24.hu, HVG, Népszava
Main items of the accounts subcontracted research gross HUF 152.4 million; communication and IT services HUF 37.3 million; property rental HUF 22.5 million (eight months); wages HUF 69.2 million; other operations HUF 6.27 million; repaid remainder HUF 12.3 million based on the Prime Minister’s Office’s statement: 24.hu, HVG, Népszava
Second grant HUF 300 million, without competition, disbursed 18 March 2026; transferred back in full 14 April 2026 24.hu, Telex, Népszava
R&D spending, Hungary 1.46% (2019), 1.63% (2021), 1.39% (2022), 1.38% (2023), 1.31% (2024) of GDP Eurostat, rd_e_gerdtot (updated 18 March 2026)
R&D spending, comparison (2024) EU-27: 2.24%; Austria: 3.26%; Czechia: 1.82%; Poland: 1.41% Eurostat, rd_e_gerdtot
Sectoral breakdown (2024, by sector of performance) Hungary: business 0.99%, higher education 0.18%, government 0.13%; EU-27: 1.49%, 0.48%, 0.24% Eurostat, rd_e_gerdtot
Government R&D target above 2% within 4 years, around 3% over 8 years Portfolio, 17 September 2026
HUN-REN funding in 2026 HUF 85 billion (according to an earlier government statement) Portfolio, 17 September 2026
HUN-REN interim president András Stipsicz asked; decision: the Governing Board and the Supervisory Board Telex, Magyar Nemzet, 17 September 2026

Three data points require a separate note. First: the items published by the Prime Minister’s Office add up to 299.97 million forints, meaning the statement covers the whole of the 2025 grant. The 152.4 million forint subcontracting item is 50.8 per cent of this. Second: HVG’s article on Vince Szalay-Bobrovniczky’s reaction gives the date of the first grant as “December 2025”. The other sources, including HVG’s first article on the complaint, give 12 February 2025; the table uses the latter. There is also a discrepancy in the number of participants in the public procurement: according to 24.hu and HVG, three companies took part and the other two bids were declared invalid, while Népszava’s MTI-based text mentions one other bid. Third: the 1.3–1.4 per cent mentioned by the minister is consistent with Eurostat’s 1.31 per cent for 2024. Eurostat’s sectoral data show the sector performing the research, not the one funding it; the public sector’s gap therefore indicates the capacity of research institutions.

6.3 Policy dimensions

  • Digitalisation and AI regulation (programme points) — public funding of artificial intelligence research fits into the measurement logic of D12 (AI productivity impact analysis and strategy); because the accounted materials appear machine-generated, the transparency principle of D1 (responsible AI framework) also applies to research products made with public money;
  • Transparency and anti-corruption policy (programme points) — the individual funding decision and the related public procurement fall under A2 (public procurement transparency), the public results repository under A1 (public funds dashboard), and the governance guarantees of the research network under A6 (checks and balances);
  • Economy (programme points) — the path of the R&D target is part of G4 (innovation ecosystem), private capital co-financing of G12 (venture capital mobilisation), and the annual evaluation of G20 (Drucker audit). The baseline value of the G4 programme point (1.6 per cent) should be updated on the basis of Eurostat’s 2024 figure of 1.31 per cent;
  • Education (background material) — the doctoral training emphasised by the minister and the careers of young researchers, as well as the reintegration of the research institutes transferred to ELTE, lie on the boundary between higher education and the research network.

6.4 Literature in detail

6.4.1 Robert Klitgaard: Controlling Corruption

Klitgaard summarises the basic components of corruption in a stylised equation (corruption = monopoly + discretion − accountability), and justifies it as follows:

“Illicit behavior flourishes when agents have monopoly power over clients, when agents have great discretion, and when accountability of agents to the principal is weak.”

All three factors were present in the MIKK funding. The funding depended on the proposal of a single official, the rules of discretion were not public, and accountability only began after the change of government, through ex-post control. In the same chapter Klitgaard writes about the selection of agents. In one of his footnotes he cites the practice in which the US federal government screens businesses and institutions receiving financial assistance before the award. In doing so, it checks their financial background, earlier audit findings and operational problems. According to the book, a rule is not good or bad in itself: a rule that creates a monopoly is an instrument of abuse, whereas a rule that narrows discretion and makes accountability easier reduces the risk. The channel rule and the minimum eligibility requirements of proposal 3.1 are precisely such discretion-narrowing rules. A later chapter of the book, discussing incentives in the public sector, also mentions semi-autonomous agencies and public–private hybrid organisations, which encourage performance while retaining public oversight:

“A variety of public-private hybrids will be created to facilitate financial incentives and competition while preserving public oversight.”

Within this framework, the planned hybrid governance of HUN-REN works if both autonomy and public oversight are laid down in law. That is why, under 3.3, MIAK asks for the board majority rule and periodic public evaluation.

📖 Source: Robert Klitgaard: Controlling Corruption

6.4.2 Ha-Joon Chang: 23 Things They Don’t Tell You About Capitalism

In the chapter on the planning elements of capitalist economies, Chang shows that through research funding the state also shapes the direction of technological development (quoted here in English translation from the Hungarian edition):

“Moreover, in all capitalist economies, by funding a very high proportion (20–50%) of research and development, the government is also planning the future of the nation’s technological development.”

According to the book, between the 1950s and the 1980s state funding of R&D in the United States was 47–65 per cent, depending on the year. Most of the sectors that gave the US its technological edge (computers, semiconductors, aircraft, pharmaceuticals, biotechnology) received significant federal R&D support through military or health programmes. Applied to the Hungarian debate: the 2 per cent target cannot be met through business incentives alone. According to Eurostat data, it is precisely public-sector research that lags furthest behind the EU average, so the path must also lay down the state share. Chang’s argument, however, cuts both ways: if the state plans the technological future, then the quality of the planning is decisive. Research funding awarded without a public call, on a single proposal, is exactly the opposite of what the successful examples show. That is why MIAK links the path under 3.2 to the channel rule under 3.1.

📖 Source: Ha-Joon Chang: 23 Things They Don’t Tell You About Capitalism (Hungarian edition: 23 dolog, amit nem mondtak el a kapitalizmusról)

6.4.3 Peter F. Drucker: The Effective Executive

Among the elements of an effective decision, Drucker names feedback last:

“Finally, a feedback has to be built into the decision to provide a continuous testing, against actual events, of the expectations that underlie the decision.”

The 2 per cent R&D target is a spending target. If it is not linked to an expectation laid down in advance of what it should deliver, there is nothing against which to measure the actual result. The five output indicators of proposal 3.2 build this feedback into the decision. In the same book, however, Drucker also warns of the limits of measurement. In the chapter on priorities he writes that research performance depends more on the courage to seize opportunities than on ability:

“Those research scientists who pick their projects according to the greatest likelihood of quick success rather than according to the challenge of the problem are unlikely to achieve distinction.”

From this follows the other half of MIAK’s proposal. The output indicators should be evaluated not project by project and annually, but at portfolio level, over a five-year window. In this way measurement does not steer researchers towards safe, short projects. As the first rule of priorities, Drucker recommends choosing the future over the past. The minister put it similarly: he would close the review of past cases next month, and from then on focus on the future. For this future, the rules of funding provide the framework.

📖 Source: Peter F. Drucker: The Effective Executive

6.5 International comparison

The established European models of state research funding rest on a common principle: the government sets the framework and the amount, but decisions on individual research funds are taken by an independent professional body. In the United Kingdom this principle has been called the Haldane principle since the early 20th century, and it is also laid down in the law on research funding. In Germany, the German Research Foundation (DFG), as a self-governing organisation with elected review panels, distributes federal and state money. The institutes of the Max Planck Society operate in a self-governing manner alongside state funding. The annual increase in the budgets of research organisations is committed to by the federal government and the states in a pact for research and innovation, so funding is predictable several years ahead. In the calls of the European Research Council (ERC), international review panels decide solely on the basis of scientific excellence. Under the rules of the EU research framework programme, the substantive part of the project must be carried out by the beneficiary, and subcontractors may only be involved for limited sub-tasks.

The Hungarian situation differs from these on two points. First, according to the MIKK case, alongside the channel built on independent review (NKFIH’s calls), research money could also flow through other funds, by individual decision. Second, the governance of the research network has changed several times since 2019, without statutory guarantees of stable self-governance. The trade union of the Hungarian Research Centre for Linguistics writes of continuous restructuring now lasting eight years. Based on international experience, the hybrid model can be lasting if the board majority, multi-year funding and independent evaluation are laid down in law, rather than depending on the goodwill of any particular government.

Digitalisation and AI regulation

  • D1 — Responsible AI framework
  • D12 — AI productivity impact analysis and strategy

Transparency and anti-corruption policy

  • A1 — Public funds dashboard
  • A2 — Public procurement transparency
  • A6 — Strengthening checks and balances

Economy

  • G4 — Innovation ecosystem and active development policy
  • G12 — Venture capital mobilisation
  • G20 — Economic policy impact assessment system (Drucker audit)

Proposed new programme point: Channel rule for public research money (research-funding body only, open call for proposals, independent review, subcontracting cap, public results repository) — for the Transparency and anti-corruption policy area.

6.7 List of sources

Press sources (MIAK press monitor, 18 September 2026 — topic 3):

Knowledge-base references (books):

  • 📖 Robert Klitgaard: Controlling Corruption
  • 📖 Ha-Joon Chang: 23 Things They Don’t Tell You About Capitalism (Hungarian edition: 23 dolog, amit nem mondtak el a kapitalizmusról)
  • 📖 Peter F. Drucker: The Effective Executive

MIAK internal materials:

  • MIAK policy area: Digitalisation and AI regulation (programme points; programme point ID: D1, D12)
  • MIAK policy area: Transparency and anti-corruption policy (programme points; programme point ID: A1, A2, A6)
  • MIAK policy area: Economy (programme points; programme point ID: G4, G12, G20)
  • MIAK policy area: Education (background material)
  • MIAK press monitor, 18 September 2026 — topic 3, score: 87/100
  • MIAK press monitor, 19 September 2026 — topic 6 (supplementary source on researchers’ pay situation)

Supplementary public data sources:

  • Eurostat — Gross domestic expenditure on R&D (GERD) by sector of performance (rd_e_gerdtot), percentage of GDP, 2019–2024, updated 18 March 2026

Generation metadata