Part I — Situation overview

On 17 August 2026 Politico Europe published internal documents centred on a confidential letter dated 4 June 2026. The letter was sent by the Egyptian Nuclear Power Plants Authority (NPPA) — the country’s authority responsible for nuclear power plants — to a Rosatom executive about the construction of the El Dabaa nuclear power plant. The document refers to “defects” affecting several reactor units, to the violation of “nuclear safety culture” and to the “deliberate negligence” of a project manager. The technical term nuclear safety culture comes from the conceptual set of the International Atomic Energy Agency (IAEA), and denotes that organisational practice in which the safety consideration comes before all others — deadline, cost, prestige — and in which the reporting of mistakes is followed not by sanction but by analysis. The letter also lists concrete items: concrete structure defects behind the metal cladding of the fourth unit, defects appearing in the base plate of the first three units that were repaired over the course of a year, and it also refers to a serious construction accident whose injured party — deliberately, according to the letter, in order to cover up the incident — was transported not by ambulance but by private car.

The origin and the limits of the factual material have to be recorded just as precisely as its content. According to the paper’s own statement the documents were passed to it by an intelligence official who asked for anonymity. In its written statement Rosatom claims that it maintained the highest level of nuclear safety throughout, and puts it that none of the accusations was confirmed by the project’s contractual, technical and supervisory procedures; it did not answer the question of whether it had received the Egyptian authority’s letter. The NPPA did not react to the paper’s repeated approaches. According to a spokesperson for the European Commission the body was not aware of the documents, and recalled that questions of nuclear safety fall within the responsibility of the member state concerned. Four further internal Rosatom documents, according to the paper, are about delays, construction quality problems and project management deficiencies; a 2025 internal audit draft writes of a significant risk to the fulfilment of the contractual obligations, and reckons that the preliminary schedule of the first unit may slip by eighteen months — from September 2028 to March 2030.

The Hungarian relevance is not indirect. The El Dabaa project uses the same reactor technology as the Paks II investment, whose main contractor is likewise Rosatom, and whose construction works began in February 2026. Over the summer the Hungarian government started a review of the investment, primarily because of the budgetary order of magnitude and the Russian tie. MIAK’s reading is therefore the following: today’s material is not about there being a defect at Paks II — but about the fact that the authority of another state working with identical technology and ordering from the same main contractor has signalled problems in official form. For the Hungarian institutional system this is not a political question but one of nuclear energy administration: for the Hungarian Atomic Energy Authority (OAH), as a body with autonomous competence independent of governmental instruction, a regulatory signal documented on an identical reactor type is a starting point for examination. And the character of the problem is that the Hungarian public today learns from the source-protected material of a foreign paper information which its own institutional system ought to be publishing regularly.

Part II — Foundations in the literature

Three authors provide the interpretative frame of the topic, and none of them writes about nuclear energy — all three write about how a large organisation deals with bad news. Daniel Kahneman, the Israeli-American psychologist awarded the Nobel memorial prize in economics for his research into decision-making biases, introduces the concept of the planning fallacy in his book Thinking, Fast and Slow: estimates given for the outcome of plans stand systematically close to the best case, not to the realistic situation. The only reliable way of correcting this is the outside view, that is, relying on the actual outcome statistics of similar projects. This proposition is the key to judging the eighteen-month schedule slippage. The economists Carmen M. Reinhart and Kenneth S. Rogoff, in their work This Time Is Different, which processes eight centuries of financial crises, show that decision-makers and investors recurrently fall into the illusion that their own case is an exception — the warning signs are not absent, they are merely set aside. The American investor Ray Dalio, in his volume Principles, names the identification of problems and the principle of zero tolerance towards them as the basic condition of organisational functioning, and introduces the concept of believability-weighted decision-making — this is the source of MIAK’s G19 programme point. The detailed treatment of the literature — author by author, with quotations — can be found in the 6.4 Literature in detail section.

Part III — MIAK’s concrete proposal

MIAK proposes three measurable measures. None of them takes a position on the authenticity of the leaked documents — all three are about what the Hungarian public administration has to do in such a situation.

3.1 A regulatory enquiry to the Egyptian authority and to the IAEA, with a public answer (within 30 days)

MIAK proposes that within thirty days the Hungarian Atomic Energy Authority should turn through official channels to the Egyptian Nuclear Power Plants Authority and to the International Atomic Energy Agency with the question of whether any finding concerning construction or safety culture has been observed on the VVER-1200 reactor type and in the main contractor’s organisation of work which is relevant from the point of view of the Hungarian licensing procedure. The authority should publish the fact of the enquiry and — omitting any classified parts — a summary of the answer received, even if the answer is negative or fails to arrive. This step is important because the regulatory channel is the only one which can produce verified information: MIAK cannot and does not wish to judge the authenticity of a source-protected package of documents, whereas an enquiry between authorities is an instrument suited to this. The step follows the logic of the KP16 information advantage programme point: what has to be done is not to react to the news but to make the facts behind the news verifiable. As a supplement it is worth asking the World Association of Nuclear Operators (WANO) as well for a summary of the shareable operating experience relating to the type.

3.2 Making the Paks II risk register public, updated annually (from the fourth quarter of 2026)

An investment of the order of ten billion euros, financed from a loan, has a risk register — the only question is who sees it. MIAK proposes that from the fourth quarter of 2026 a public version of the investment’s risk register should appear annually, and should contain at least: (a) the current state of the schedule by unit, with the deviation measured against the earlier plan; (b) a summary of the contractual penalty and guarantee system; (c) the documented performance shown by the main contractor on other projects — the Egyptian case now come to light would be precisely such an item; (d) the exposure of the sanctions and supplier chain; (e) the financing and exchange rate risk. The G19 radical transparency programme point prescribes this logic generally as well, and the G20 impact assessment system prescribes subsequent evaluation. In Dalio’s frame (see 6.4.3) the register does not express the absence of trust but the condition of trust: making problems visible is the only way of getting them solved. The A2 public procurement transparency and A8 accountability programme points ask for the same level of publicity on the side of the use of funds.

3.3 A quantified exit scenario, regardless of the direction of the decision (by the first quarter of 2027)

MIAK proposes that by the first quarter of 2027 the exit scenario of the Paks II investment should be prepared, and that its quantified summary should be made public — regardless of whether the government decides to continue or to stop the project. The document should contain: how large is the sum already paid out and the sum recoverable under the contract; how much the contractual penalty is and on what conditions it can be enforced; what alternative technology is available and with what lead time; and what the electricity balance up to 2035 would be without this capacity. This is the direct application of the KP22 exit strategy planning protocol programme point, which prescribes an obligatory exit memorandum for every long-term commitment exceeding one billion euros. The programme point itself treats the document as an internal material, because it may be a bad signal towards the negotiating partner; MIAK therefore proposes here the publicity not of the full material but of the quantified summary. The aim is not to question the project but to strengthen the negotiating position: whoever knows what exit costs can also bargain for better conditions for staying. The capacity-side consequences have to be modelled within the framework of the K2 energy transition and K7 shock resilience programme points, and the alliance and geopolitical aspect is given by KP11 strategic balance policy and HV12 geostrategic defence planning.

The common principle of the three proposals is that none of them demands a decision about the future of the investment. All three do the same thing: they produce information ahead of the decision. In Kahneman’s frame (see 6.4.1) this is the institutionalisation of the outside view — the question is not what the decision-maker thinks about their own project but how similar projects have performed. And in Reinhart and Rogoff’s frame it is the prevention of the “this time is different” reflex: a warning sign is worth something if there is a procedure that obligatorily records and answers it.

Part IV — Expected effects and risks

Dimension Expected effect Risk
Economy The risks of the investment are built into budgetary planning in quantified form; the financing premium may fall The publicity of the risk register may trigger a contractual dispute if it touches an element qualifying as a business secret
Energy supply The plan for replacing baseload capacity is prepared in advance, there is no need to improvise in a crisis The public summary of the exit scenario may act as a self-fulfilling signal on the supplier market
Public administration The regulatory enquiry creates a precedent: the authority obtains international information rather than waiting for press reports If no substantive answer arrives to the enquiry, the public will see the authority as powerless
Foreign policy The Hungarian position will be procedural and not ideological, which is defensible in both directions The topic may slide into an alliance dispute where the procedural argument is read as a political position

The main trade-off of the package of proposals lies between publicity and the negotiating position. A public risk register and a published exit summary may in principle weaken the Hungarian side’s bargaining position, because they show how large the price of exit is. In practice, however, the partner knows this price anyway — it has the contract — while the Hungarian public and the National Assembly do not. The information asymmetry therefore stands today to the detriment not of the partner but of the domestic public. The proposal tips over to the risk side if the material made public goes into technical details which have to be handled confidentially for safety reasons: this is why MIAK proposes summary-level, indicator-based publication, with itemised reasons for redaction. The second fault line runs at the regulatory enquiry: the independence of the OAH also means that initiating the enquiry cannot be the subject of a governmental instruction — MIAK therefore does not propose an instruction but that the authority should act within its own competence and inform the public about it.

Part V — Measurability and summary

5.1 What is worth following? (proposed KPIs)

Four proposed performance indicators (KPIs) are worth watching:

  • Regulatory response time: by the end of the fourth quarter of 2026 a public summary should appear on whether a substantive answer has arrived from the Egyptian authority and from the IAEA, and what its content is.
  • Existence of the risk register: by the end of 2026 the public risk register of Paks II should be available, including the schedule deviation by unit; from 2027 updated annually.
  • Measurement of schedule deviation: the difference between the original and the current commissioning date expressed in months, published every six months — this is the indicator which makes the effect of the planning fallacy visible.
  • Exit scenario: by the end of the first quarter of 2027 the quantified summary should be available on four items — sum paid out, recoverable sum, value of the contractual penalty, replacement capacity requirement up to 2035.

5.2 Summary

MIAK’s key message is that the correct answer to a leaked package of documents is not to believe or to reject it but to verify it through regulatory channels. Of the decision-maker MIAK asks a single, cheap and quickly feasible step: an enquiry by the Hungarian Atomic Energy Authority to the Egyptian authority and to the IAEA within thirty days, publicly confirmed. This takes a position neither for nor against the investment — it merely obtains information, where it can be obtained. And of the public it asks that in the coming weeks it should ask not whether the accusations are true, but this: has anyone asked officially, and what was the answer.

Two MIAK foundational values are directly in play here. Data-drivenness: an investment of the order of ten billion euros cannot be decided on the basis of press reports — neither its continuation nor its stopping — the decision requires a risk register, schedule data and a quantified exit calculation. And transparency: if the Hungarian public learns about the risks of its own largest investment from a foreign paper, then it is not the paper that is over-performing but the domestic system of publicity that is under-performing. The risk register is not a gesture of distrust towards the contractor but the minimum without which the National Assembly cannot perform its task of budgetary control.

Part VI — Reasoning and further sources

6.1 The press framing by spectrum

The topic on this day is international in origin, and this also determines its framing. Politico Europe, as an EU policy paper, ran the story on two threads: the content of the Egyptian regulatory letter and the EU consequence. The latter was formulated most precisely by the highlight placed under the headline, when it recorded that the Egyptian plant uses the same technology as the project in Hungary. The paper carefully marked the uncertainty throughout: the anonymity of the source, Rosatom’s denial, and the fact that the Egyptian authority did not react to the approaches. This framing is professionally correct, but for the Hungarian reader it lacks what only a domestic source can provide: what the Hungarian authority knows and what it has done so far.

In the Hungarian press this thread remained in the shadow of the Paks cooling water crisis in the news flow at the beginning of the week. The domestic papers reported predominantly on the level of the Danube and on the construction of the bottom sill, and the Rosatom thread appeared at most in connection with the prime minister’s earlier statement questioning the contract. This shift of emphasis is understandable — the acute security of supply risk is closer — but it has a structural consequence: the short-term water crisis crowds the long-term investment risk off the agenda, although the two concern the same power plant. MIAK therefore handles the two topics on separate threads, and alongside the analysis of the institutional management of the Danube as a shared resource of 16 August 2026, carries the main contractor and regulatory supervision axis further here.

A further framing consideration arrived from the British defence and security expert band: the analysis of the Royal United Services Institute was about the risks of business relations with Rosatom, referring to intelligence ties and to the role around the occupied Zaporizhzhia plant. This framing gives the security policy reading, and is an important counterpoint to the purely technical approach — at the same time, for MIAK it does not replace regulatory fact-checking: the analysis of an expert institute is not a regulatory finding.

6.2 Facts and data

Data Value Source
Date of the Egyptian regulatory letter 4 June 2026 Politico Europe
Publication of the documents 17 August 2026 Politico Europe
Project concerned El Dabaa nuclear power plant, 4 units, Egypt Politico Europe
Reactor type VVER-1200 — identical to the one planned in the Paks II project Politico Europe
The Egyptian prime minister’s July schedule all four units by 2030, the first unit producing from 2028 Politico Europe
Slippage estimated in the internal audit draft for the first unit 18 months (September 2028 → March 2030) 2025 internal document cited by Politico Europe
Nuclear fuel on site not yet delivered Politico Europe
Start of the Paks II construction works February 2026 Politico Europe
The European Commission’s position was not aware of the documents; nuclear safety is a member state responsibility spokesperson for the European Commission (Politico Europe)
Rosatom’s position maintained the highest level of safety; the accusations were not confirmed by the project’s procedures Rosatom statement (Politico Europe)

The most telling row of the table is the eighteen-month slippage estimate. Not because it would be dramatic in itself — with megaprojects this order of magnitude is usual — but because it comes from an internal document, that is, the organisation already saw its own schedule as untenable in 2025, while a different date appeared outwardly. The Hungarian lesson from this is not Egyptian: the question is whether in the Paks II investment there exists a procedure which obligatorily reflects the internal risk estimate in the public schedule as well. Today there is no public answer to this.

6.3 Policy dimensions

  • Environment and climate (programme points) — the relationship of baseload capacity and the decarbonisation path is the subject of the K2 energy transition; the management of capacity loss belongs to the K7 shock resilience programme point.
  • Economy (programme points) — the publicity of the risk register is the direct application of G19 radical transparency; the obligation of subsequent evaluation belongs to the G20 impact assessment system.
  • Foreign policy (programme points) — the exit scenario is an obligatory element under the KP22 exit protocol; regulatory information gathering is the logic of KP16 information advantage; the management of supplier dependence is the frame of KP11 strategic balance policy.
  • Transparency and anti-corruption policy (programme points) — contractual and procurement publicity is the subject of A2, the accountability of the use of funds that of the A8 programme point.
  • Defence (programme points) — the supplier exposure of critical energy infrastructure belongs on the risk map of HV12 geostrategic defence planning.

6.4 Literature in detail

6.4.1 Daniel Kahneman: Thinking, Fast and Slow

Kahneman derives the concept of the planning fallacy from a failure of his own: the completion of a textbook-writing project was estimated at two years, and the work in the end took far longer. The lesson consists of three parts, and all three are directly applicable to megaprojects:

“The second lesson was that our original forecasts of about two years for the completion of the project had suffered from a planning fallacy. Our estimates were closer to a best-case scenario than to a realistic assessment. I was slower to accept the third lesson, which I call irrational perseverance: the folly we displayed that day in failing to abandon the project.”

For him the solution is not good intentions but procedural: the organisation has to institutionalise the outside view, that is, it has to measure its own plan against the actual outcome statistics of similar projects. Kahneman names concrete organisational instruments for this — checklists, reference class forecasting and the premortem analysis, in which one has to formulate before the decision why the plan will fail. The eighteen-month internal slippage estimate around the El Dabaa project is precisely the signal that the outside view would put to use: it is not interesting in itself but as a basis of comparison, alongside the schedule performance of other VVER constructions. This is why MIAK’s proposal 3.2 asks for the regular, public publication of the schedule deviation by unit — this is the indicator which makes organisational optimism measurable.

📖 Source: Daniel Kahneman: Thinking, Fast and Slow

6.4.2 Reinhart and Rogoff: This Time Is Different

Reinhart and Rogoff processed eight centuries of financial crises, and summed up the pattern in the volume’s title phrase:

“Major default episodes are typically spaced some years (or decades) apart, creating an illusion that ’this time is different’ among policymakers and investors.”

The authors’ argument is not that the actors are stupid. The argument is that the time elapsed between individual cases is long enough for institutional memory to wear away, and for every new actor to see their own situation as exceptional. The megaproject analogy is close: nuclear constructions are rare, a given country builds one or two in a generation, so there is no continuous practice that would automatically pass on the earlier experience. From this follows one element of MIAK’s proposal: the performance shown by the main contractor on other projects should not be informal background knowledge but an itemised row of the risk register. The Egyptian case now come to light would be precisely such an item — regardless of whether the accusations eventually prove well founded, the fact of the regulatory signal is in itself information which a risk register has to record.

📖 Source: Carmen M. Reinhart and Kenneth S. Rogoff: This Time Is Different — Eight Centuries of Financial Folly

6.4.3 Ray Dalio: Principles

Dalio locates one of the basic principles of organisational functioning in the relationship to problems: for him the second big step after setting goals is “identifying problems and not tolerating them”. To this he attaches the idea that for a good decision one’s own point of view is not sufficient:

“Those who care about making the best decisions rarely trust that they already have the best answer. That is why they seek to find out more — often by learning about the thinking of other believable people, especially those who disagree with them.”

This is the principle of believability-weighted decision-making, which is the source of MIAK’s G19 programme point: not every opinion is worth the same, but neither is the decision-maker’s own opinion privileged — the weight is given by proven performance. In the case of Paks II this has two practical consequences. One is that counter-arguments have to be documented, not deflected: the public reasoning about the investment has to contain what speaks against it too, and what the answer to that is. The other is that the main contractor’s own statement — in either direction — is not sufficient evidence: Rosatom’s denial is just as much the claim of an interested party as the Egyptian authority’s letter. This is why MIAK proposes an independent answer obtained through regulatory channels, and not the weighing of statements.

📖 Source: Ray Dalio: Principles

6.5 International comparison

Three European practices offer a usable model. The Finnish case is the most instructive: the construction of the Olkiluoto 3 unit slipped by thirteen years, and during the process the Finnish radiation safety authority (STUK) regularly published its supervisory findings, including the deficiencies of supplier quality assurance. This practice did not speed up the construction, but it preserved the authority’s credibility — the public learned even the bad news from the authority, not from the press. The British model gives an example from the investment side: the HM Treasury Green Book makes the evaluation of the exit option obligatory for multi-billion contracts, which is the direct forerunner of MIAK’s KP22 programme point. And the Slovak and Czech experience is relevant on the question of supplier diversification: both countries operate units of Russian technology, and switching fuel supply to a Western supplier was a multi-year, documented process — that is, technological dependence can be loosened, but not quickly and not for free.

The EU level is deliberately restrained on this question: the nuclear safety directive places responsibility on the member state authority, and the European Commission’s spokesperson confirmed this in the present case too. The practical consequence of this is that in a Hungarian matter a Hungarian authority has to act — one cannot wait for an EU investigation. This strengthens the weight of MIAK’s proposal 3.1: if the OAH does not ask, nobody will.

Environment and climate

  • K2 — Energy transition plan
  • K7 — Energy market shock resilience

Economy

  • G19 — Radical transparency in economic decision-making
  • G20 — Economic policy impact assessment system

Foreign policy

  • KP11 — Strategic balance policy
  • KP16 — Foreign policy information advantage strategy
  • KP22 — Exit strategy planning protocol

Transparency and anti-corruption policy

  • A2 — Public procurement transparency
  • A8 — Cohesion policy accountability

Defence

  • HV12 — Geostrategic defence planning

Proposed new programme point: An obligatory, public risk register for every state megaproject exceeding one billion euros, including the performance shown by the main contractor on other projects — for the Economy area.

6.7 List of sources

Press sources (MIAK foreign press monitor, 17 August 2026 — topic 1):

  • [Politico Europe] ‘Deliberate negligence’: Russian nuclear power company with EU operations accused of violating safety standards — https://www.politico.eu/article/rosatom-hungary-nuclear-energy-egypt-russia-el-dabaa-paks-ii-russia-safety-standards-violations/
  • [Deutsche Welle] Romania shuts nuclear plant, Hungary dams dry Danube River — https://www.dw.com/en/romania-shuts-nuclear-plant-hungary-dams-dry-danube-river/a-78359581
  • [Deutsche Welle] When extreme heat threatens Europe’s nuclear power — https://www.dw.com/en/when-extreme-heat-threatens-europe-s-nuclear-power/a-78297778
  • [Euractiv] Spain delays nuclear power plant shutdown — https://www.euractiv.com/news/spain-delays-nuclear-power-plant-shutdown/

Knowledge base references (professional literature):

  • 📖 Daniel Kahneman: Thinking, Fast and Slow
  • 📖 Carmen M. Reinhart and Kenneth S. Rogoff: This Time Is Different — Eight Centuries of Financial Folly
  • 📖 Ray Dalio: Principles

Note: the local file path of the books does not appear in the visible text of the blog — only the author and the title. The file path is an internal matter of the generation process, not the reader’s.

MIAK internal materials:

  • MIAK policy area: Economy (programme points; programme point ID: G19, G20)
  • MIAK policy area: Foreign policy (programme points; programme point ID: KP11, KP16, KP22)
  • MIAK policy area: Environment and climate (programme points; programme point ID: K2, K7)
  • MIAK policy area: Transparency and anti-corruption policy (programme points; programme point ID: A2, A8)
  • MIAK foreign press monitor, 17 August 2026 — topic 1, score: 94/100

Supplementary public data sources (where used):

  • Hungarian Atomic Energy Authority — licensing decisions and annual reports
  • International Atomic Energy Agency (IAEA) — PRIS database and operating experience reports
  • WANO — public peer review summaries
  • MAVIR — supply-side capacity analysis
  • ENTSO-E — Ten-Year Network Development Plan

Generation metadata