Two steps takenFive steps to be taken

In June 2026, Romania faces a double diagnosis delivered by two independent entities, guided by two different methodologies, reaching the same conclusion. Edge Institute and Digital Nation, in the „Digital Governance Framework”, a report published in October 2025, give Romania a score of 3.1 out of 10 for „digital state maturity”. The OECD, in the „Digital Government Index 2025”, a study published in February 2026, gives Romania a score of 0.24 out of 1.0, nearly one third of the OECD average (0.70) and less than a quarter of the score of South Korea, the ranking leader (0.95).

„AI will not reform the state's operating model. It will scale its dysfunctionality faster than it will deliver reform.”

Ott Velsberg, former Chief Data Officer of the Estonian Government (2018–2025), in AI Won’t Transform Government Without Governable Data, Tony Blair Institute, 6 May 2026.

In June 2025, Edge Institute launched, together with other organisations, the „Call for Romania's Digital Transformation”.

In October 2025, together with Digital Nation, we published the „Digital Governance Framework in Romania”, a comprehensive diagnostic of digital state maturity, assessed at 3.1 out of 10.

As a fundamental solution, the report proposed a four-layer governance model: a Deputy Prime Minister for Digital Transformation, a CIO (Chief Information Officer) within the central Government Working Apparatus, a Digital Council that adopts binding decisions, and an implementation agency, including through the reform of ADR and the transfer of the institution under the authority and leadership of the GCIO, with full execution capacity.

In November 2025, we organised, under the High Patronage of the President of Romania, the Digital Governance Summit, hosted at Cotroceni Palace, opened by H.E. Mr Nicușor Dan and supported through the interventions of five ministers overseeing portfolios relevant to Romania's digitalisation. The event will be held again, in a second edition, in November of this year, also at Cotroceni Palace.

Edge Institute initiatives presented above reflect three distinct and symbolic types of action dedicated to digital governance, in addition to the dozens of articles and studies published on the same topic over the past year.

Today, in May 2026, two steps have been taken. Five remain to be taken.

The completed steps are:

• the appointment of a Deputy Prime Minister with responsibilities in the sphere of digital transformation (appointed in November 2025 with a mandate for digitalisation and debureaucratisation);
• the launch of a conversation about the reform and restructuring of the Authority for Romania's Digitalisation.

Although both steps represent undeniable progress, their impact is insufficient without continuity and, above all, without an acceleration of the subsequent measures presented below.

The economy is stagnating: GDP fell by 1.7% in the first quarter of 2026 and the full-year projection stands at 0.1%. Concordia Employers' Confederation estimates that coherent and coordinated digitalisation can generate annual growth of 1–1.5% of GDP. Without the governance architecture proposed in this document, Romania will negotiate the 2028–2034 Multiannual Financial Framework with the same capacity deficit it brought into the current cycle. And, in all likelihood, will deliver the same results.

„Inflation will be defeated through the acceleration of digitalisation and investment in AI.”

Mario Draghi, 17 May 2026 (source)

The five steps we consider essential in the new government's programme:

1. 1.

   Reconfirming the position of Deputy Prime Minister for Digital Transformation, with a clear mandate dedicated to accelerating the digitalisation of public services.

2. 2.

   Appointing the ADR President as Government CIO, with the rank of State Secretary (a rank already carried by the ADR President position), at the helm of a reformed ADR, placed at the Centre of Government, apolitical, with a clearly defined technical mandate and a solid organisational capacity in its reformed structure.

3. 3.

   Operationalising the Digital Council through the restructuring of CERB and the expansion of the role and structure of this body's Consortium, through a monthly calendar of mandatory meetings and the adoption of binding, executive decisions, along with the necessary legislation to operationalise this measure.

4. 4.

   Launching a dedicated team for the Government Digital Wallet: a dedicated unit of at least 20 specialists with a single leader, inspired by the Swedish model, a model that will also be illustrated in an Edge Institute study following exchanges of best practices conducted with the Swedish Agency for Digital Government, as a first pilot of the model. The success of this working model should form the basis for reorganising ADR from an intermediary between European funds and public administration into an orchestrator of public service digitalisation, through the creation of a Digital Delivery Unit, building on the dedicated government digital wallet team model, to act as an implementation structure for digital initiatives and reforms across central public administration. The precedent exists in Romania as well: a delivery unit-type team, with international expertise, operated at Palatul Victoria in 2014 and contributed to the launch of ANAF's Virtual Private Space (SPV).

5. 5.

   Building governance and artificial intelligence utilisation capacity. AI is already widely used without knowing the actual adoption rate in public administration or education, without guidelines, without a framework, without consistent operational capacity. Five deliverables: a mandatory AI usage guide for public administration adopted by Government Decision, an AI product procurement framework integrated into ANAP procedures, at least two pilot programmes in public services with public reporting and efficiency indicators, the inauguration of the national AI regulatory sandbox before August 2027, and a national AI literacy programme for public administration. Compliance with EU obligations, implementing legislation, ANCOM becoming operational, Digital Omnibus deadlines, remains the mandatory minimum threshold.

In parallel, two points addressed to the Presidential Administration as a layer of continuity in a country with frequent government changes, following the recent appointment of a State Counsellor for AI: adapting the diplomatic network to the digital transformation objective and launching a „National Compass for Digital Diplomacy”, as well as launching a cross-party National Pact, starting from the premise of the national digital consensus reflected in all surveys conducted by Edge Institute. These clearly show that, although Romanians display a consistently growing digital optimism, rising from 72% to over 78% between November 2025 and March 2026 alone (AtlasIntel study for Edge Institute), the perception of the slowness of the state's digitalisation remains at a similarly high level of around 84% from one cycle to the next.

Thus, this cross-party document brings the importance of governance back to the attention of the incoming Government, in a context where, beyond Edge Institute's actions over the past year, there are numerous public voices, from central public administration, civil society, and the private sector, who share the same conclusion.

Without clear leadership, we cannot accelerate a command centre close to the Prime Minister, nor an efficient decision-making and implementation structure.

We will draw on two concrete case studies (the Government Digital Wallet and Building AI Governance and Utilisation Capacity) to illustrate the bottlenecks caused by the absence of governance. The first represents Romania's most important project related to the digitalisation of public services, the government wallet, while the second concerns the way in which the state chooses to participate in the most important technological transformation of this century, namely the AI era.

Regarding the Digital Wallet, this document will be followed by a separate operational report, „Delivering the EUDI Wallet in Romania: a governance and delivery proposal” (Edge Institute, June 2026), which details the team structure, roles and competencies, built on an analysis of how other European countries are implementing the EUDI Wallet.

One of the conclusions already identified in our international analysis is that no country has skipped the step of having a dedicated team and a project leader with formal authority. So far, Romania has not fulfilled even that.

The Call from June 2025 proposed four measures addressed to the Presidential Administration: a Presidential Adviser for Digital Transformation, a Presidential Commission on the same topic, a Diplomatic Network for Digital Transformation, and a National Pact for Romania's Digitalisation.

The Digital Governance Framework, published in October 2025 on the basis of 18 institutional interviews and with Estonia as its reference point, proposed a four-layer model for the government apparatus. The Framework's diagnosis: „A vacuum in strategic digital governance.” Calculated score: 3.1 out of 10.

Neither of these two documents was ignored. In November 2025, the Prime Minister appointed a Deputy Prime Minister with a role in digitalisation and debureaucratisation, the first layer of the „Framework” to become a government position. CERB, the committee equivalent to the „Digital Council” from the Digital Nation report, a body that previously lacked a well-defined strategic agenda, began meeting more frequently and with an agenda built around fara-hartie.ro and proposals gathered from citizens.

ADR acquired new leadership with a profile suited to the mandate of developing citizen-facing products. Radu Puchiu, e-governance expert at the Aspen Institute, points out that

„the appointment does not resolve the underlying problem, ADR's role within the institutional landscape and its mandate. ADR needs to be placed within the central Government Working Apparatus (and, notably, this is something the former ADR President also agreed with). This is something that should have happened a long time ago; perhaps the new leadership will take that step.”

Nevertheless, the context of appointing new ADR leadership opens the opportunity for a structured conversation about the role of this institution. There is now an opportunity for ADR to become a clearly defined delivery unit-type agency, on the basis of greater authority and execution capacity. The preconditions, however, are its placement at the Centre of Government, under the authority of the Deputy Prime Minister responsible for the portfolio, the strengthening of institutional capacity, and the appointment of the ADR President as Government Chief Information Officer, with the rank of State Secretary, a rank already carried by the ADR President position, mandated to also coordinate the reform of the institution.

Things have moved. Not enough

The Call to the Presidential Administration has just received an institutional response, the appointment of an adviser on AI and the adoption of new technologies. There are also at least three mentions worth making regarding the Cotroceni Palace agenda and its intersection with digital transformation.

The Presidential Adviser for Economic and Social Affairs has consistently raised the topic of digitalisation through his public statements, from electronic signatures to the government wallet, the digitalisation of ANAF, and remarks on the use of artificial intelligence in the public sector.

In parallel, President Nicușor Dan announced the establishment, at the Cotroceni Palace, by the summer of 2026, of a permanent department dedicated to countering disinformation and hybrid warfare, under direct presidential coordination and composed of representatives of state institutions with responsibilities in the field. An Edge Institute report dedicated to the governance of countering disinformation discusses at length institutional models relevant to inserting such a structure under the aegis of the Presidential Administration.

On 5 May 2026, the first edition of the Just.AI Forum was held at Cotroceni Palace, during which, at the President's initiative, an institutional memorandum was signed to create the first national cooperation framework for the integration of artificial intelligence into the judicial system, with reports, diagnostics, and operational priorities expected by October 2026.

The structural conclusion of our document, that Romania needs a command centre in its digital governance architecture, does not bear the Edge Institute's stamp alone. It is an idea taken up in operational language from a diagnosis that already exists at the OECD, an international organisation Romania is seeking to join. In the „Digital Government Review of Romania” (2023), the OECD wrote explicitly: „Romania needs to further strengthen the governance of digital government, as the organisation steering the digital transformation of government is not yet positioned strategically to gain necessary political support and legitimacy across the public sector.”

The OECD added another sentence that speaks volumes about the state of digitalisation in Romania, a perspective we repeat in every article: „There is a need to transition from focusing on the adoption of legal measures to focusing on implementation.”

And one more recent chart, confirming the same diagnosis. Only the methodology differs.

In February 2026, the OECD Public Governance Committee published the results of the „Digital Government Index 2025”. The OECD average score is 0.70 out of 1.0 (up 14% from 0.61 in 2023). Denmark, one of Europe's leaders in digital transformation, holds, for instance, a score of 0.80 out of 1.0. Meanwhile, France and Spain, reference countries for Romania in terms of administrative organization, gravitate within the same sphere, with scores of 0.80 and 0.75 respectively. Even within the region, the gap between Romania and its neighbors is significant: while Romania stands at 0.24 out of 1.0, at roughly a third of the OECD average, Greece's average is almost triple, Poland's and Hungary's scores are more than double those of Romania, and Bulgaria exceeds Romania's standing by almost 50%.

The World Bank classifies Serbia (0.86), Bulgaria (0.81) and Albania (0.80) in Group A of the GovTech Maturity Index 2025, very high governmental digital maturity. Ukraine, in the fourth year of war, leads the European ranking of open government data with 97%, a Trendsetter status, alongside France and Poland; Serbia is a Fast-tracker with 83-90%; Romania is a Beginner with 65-69% (Open Data Maturity Report, European Commission, 2024). The gap relative to Estonia is cited often. The one relative to neighbors, more rarely.

In the first quarter of 2026, Romania's economy contracted by 1.7% compared to the same period of the previous year and will most likely cement a 3-year cycle, 2024-2026, in which economic growth is in fact stagnation, with modest positive values and a period average below 0.5%.

In this context, digitalizing the state is no longer an aspiration. It becomes, exactly as in the case of Greece after the debt crisis of the 2010s, an instrument of administrative efficiency and economic recovery. It is a matter of will before being a matter of money.

The European Commission estimates that, by meeting the objectives of the Digital Decade, the Union could "unlock substantial economic gains, estimated at up to 1.8% of its GDP." Mario Draghi, in his September 2024 report on European competitiveness, is even more direct: "Europe has largely missed out on the internet-led digital revolution and the productivity gains it brought." Romania cannot afford to miss the second wave as well.

"Productivity in the United States has grown over the past six years. [...] you get a gap between the US and Europe of nine percentage points compared to 2019. [...] What those who know about this tell me is that it's because of digitalization. It's a simpler explanation, but it's about the enormous efficiency gains that companies actually achieve by digitalizing their workflows and processes"

Mario Draghi at Tech Talks Timișoara, on May 17, 2026.

If for Europe the 9-point productivity gap relative to the US is explained by Draghi through digitalization, for Romania the concrete figure already exists. The Concordia Employers' Confederation, in its December 2025 study "Proposals for economic recovery and growth," quantifies it directly: "Coherent, coordinated and predictable digitalization can bring an annual growth of 1–1.5% of GDP, through reducing the administrative burden, increasing efficiency and attracting private investment in technology." At a GDP of approximately 350 billion euros, that means between 3.5 and 5 billion euros per year that depend exclusively on a decision of institutional organization. And which are lost annually, through postponement, with no subsequent compensation.

Draghi also cited a few days ago in Timișoara a new OECD conclusion. "A recent study says that half of the productivity growth between now and 2030 will be due to AI adoption. [...] Today, 37% of European companies adopt AI. [...] the scale of productivity growth at companies that have adopted AI is on the order of 4%."

Nationally, the European Commission's report on Romania's 2025 Digital Decade confirms the scale of the investment: the national roadmap comprises 98 measures with a budget of 3.6 billion euros, the equivalent of 1.01% of GDP. The money is there. We have a choice to make: either to direct it as a productive investment or as an administrative expense.

The example from the Balkans: Greece

The comparison that matters here is the one with Greece, a country that went through a decade of deep, structural economic problems that today are more familiar to us than ever before in recent history. Deficit, austerity, a crisis of public trust. In July 2019, the Mitsotakis government made a strategic bet: it appointed a minister of digitalization with a clear mandate and political authority (Kyriakos Pierrakakis), but above all it treated digitalization as an instrument of administrative efficiency and economic recovery. The launch of the gov.gr platform in March 2020, two weeks after the first COVID case, accelerated what would have taken years.

In four years, Greece delivered over 1,800 digital public services, over 85 million transactions per month (February 2026), and a rise in the Digital Decade from a country below Romania to one substantially above. This was achieved by a country with a decade of economic crisis behind it, with a budget far more constrained than the one Romania can access through the 2028–2034 Multiannual Financial Framework.

The lesson for Romania in 2026 is pure arithmetic. Administrative efficiency, measured in time saved by citizens and companies, in reducing public personnel spending, in the speed of absorbing European funds, is, for a country in recession, the measure by which trust in the state and the recovery of the economy are built.

According to the survey conducted by Ipsos for Edge Institute in 2024, 47% of Romanians consider that the digital services of private companies are more efficient than those of the state. Only 11% say the state is more efficient. The gap is fourfold, in the citizen's perception. According to AtlasIntel & Edge Institute data from November 2025, 84% of Romanians say the state's current pace is slow or very slow. 75% demand major change.

The two indicators, the state-private efficiency gap and citizen frustration, are not independent. They reflect the same reality, measured from two angles.

The two sources above, Edge Institute × Digital Nation and OECD, measure the central apparatus. For the citizen, however, the digital state begins at city hall. At that level, the diagnosis was measured by a third report.

At the Digital Governance Summit organized by Edge Institute at Cotroceni Palace in November 2025, President Nicușor Dan formulated the principle:

"Digitalization means time gained, by people and by the economy. The more you shorten the time citizens spend in interactions with the administration, the more you gain their trust in the state."

Nicușor Dan, President of Romania, November 2025.

A month later, in December 2025, the Dela0.ro journalists published, within the DigITup project run by the Center for Independent Journalism with the support of Edge Institute, the first interactive map of the digitalization of local administrations in Romania, built from 325 requests for access to public information sent to all urban city halls.

The measured verdict: national average score 4.57 out of 10. Only 22 large city halls have a digitalization strategy approved by the local council. 32 city halls still use Windows XP or Windows 7. A quarter offer strictly online tax payment, with no other electronic services. 10% of the city halls queried, including Sibiu and most of Bucharest's districts, did not even respond at all to the journalists' requests, violating Law 544/2001 on access to public information.

The president's principle remains, for now, only an aspiration.

Also in April 2026, the POLITICO Research & Analysis Division published the European Digital Capacity Index, the comparative assessment of the 27 EU states on digital capacity. Romania: 26th out of 27, overall score 51.5/100. Bulgaria closes the ranking. More telling than the rank is the structure: Governance Level (legislative framework, transposition, national strategies), 64.1, below the EU average. Ecosystem Level (infrastructure, digital skills, business adoption, public services delivered), 38.8, the weakest score in the EU.

We have the frame, but no content. On the chapter of digital skills, a dedicated subchapter, Romania scores 12 points, last in the EU. Only 9% of Romanians have digital skills above the basic level.

Four independent sources. Four levels of analysis. Four different methodologies. The same conclusion. 3.1 out of 10 at the central level. 0.24 out of 1.0 as OECD positioning. 4.57 out of 10 at city halls. 51.5 out of 100 as European positioning.

Recession paradoxically offers the clearest window for deep institutional reform. The argument "this is not the moment" becomes, in recession, exactly the opposite. This is the moment. There is no better one than this.

Romania's IT industry says the same thing we say

In the report "Stimulating the transition to innovation in Romania's ICT industry" published by ANIS, the domestic IT industry self-diagnoses its limits: the ICT sector contributes approximately 8% of the added value of the Romanian economy, among the highest shares in the EU, and grew by 17% annually between 2021 and 2023. However, the report concludes that the current growth model, based on outsourcing and cheap execution, "has reached its limits." Romania has approximately 220,000 IT specialists; the Czech Republic has 235,000 with a population almost three times smaller and produces an ICT output 19% higher. The difference is structural.

The report quantifies the possible leap in figures. If Romania reached proportionally the size of the Polish industry, the additional contribution would be approximately 38.8 billion euros to GDP, 294,500 jobs and 88 billion euros in cumulative fiscal revenue. Generalized AI adoption could add, separately, 14-16 billion euros annually to GDP, the equivalent of 5% of the total.

The same magnitude, from a parallel source: AI as an economic lever.

The Tony Blair Institute (TBI), together with Faculty AI, estimated in January 2026 that the responsible adoption of AI in the British public administration could generate savings of up to £200 billion over five years, or £40 billion annually in net productivity gains. Scaled proportionally to GDP (UK GDP ≈ £2,700 bn, RO GDP ≈ €350 bn), the equivalent for Romania would be on the order of 5-7 billion euros annually in productivity gains from AI in the public sector alone, converging with the ANIS-Implement estimate of 14-16 billion euros annually from AI across the entire economy. Two different methodologies, paired with two compatible magnitudes, deliver a similar operational conclusion.

The Concordia Employers' Confederation, in its December 2025 study "Proposals for economic recovery and growth," explicitly links digitalization to economic competitiveness.

"For Romania, digitalization can no longer be seen as a sum of projects, but as a structural reform of the state and the economy, which influences productivity, taxation, voluntary compliance, the relationship between citizen and administration, but also the capacity of companies to innovate and invest"

Concordia Employers' Confederation, Proposals for economic recovery and growth, 2025

The entire context above is built on messages and research constantly released by a series of organizations or political actors. From within the central administration or from the private sector, from civil society or academia.

This is what we at Edge Institute have been saying for almost a year as well. And we truly believe it is the moment for the next Government to listen to this chorus of voices.

The first case through which the diagnosis above is demonstrated concretely is the Government Digital Wallet, Romania's implementation of the EUDI Wallet.

The mandatory deadline set by the eIDAS 2 Regulation is 31 December 2026. All 27 EU member states must provide citizens with a digital wallet that allows the storage, presentation and verification of official documents within a user-controlled application.

Romania's status at the time of publishing this document: we have no operational project leader and no public deadline for this minimum step. We also have a Prime Minister's Decision (183/2026) designating the Deputy Prime Minister, MEDAT, MAI, and STS as members of the RO EUDIW Commission, with a mandate to ensure the institutional, legal, and technical framework for the application of eIDAS 2 provisions, and the possibility of consulting other stakeholders as needed, including from the private sector. We have a memorandum signed with Mastercard, non-exclusive, in which Mastercard is ready to provide assistance to the project.

Why it has not started: there are three key teams, those belonging to the Special Telecommunications Service (STS), the Ministry of Internal Affairs (MAI), and the Authority for Romania's Digitalisation (ADR), and none of them holds the mandate to coordinate the project. Three teams, each with different resources and expertise, and no leader. The key diagnosis of the Digital Nation report, applied to a concrete project.

In parallel, the electronic identity card (CEI), the essential pillar of the foundational infrastructure on which the Wallet should be built, has an issuance rate that places the NRRP (National Recovery and Resilience Plan) target (5 million by 30 June 2026, subsequently renegotiated) firmly in the zone of non-delivery. The risk of a penalty materialises as a pro-rata reduction of the NRRP payment tranche for the digital component, estimated by MAI at up to 264 million euros in the scenario of full non-fulfilment of the target.

This is, in project form, the complete diagnosis. The money exists, through NRRP (National Recovery and Resilience Plan) funding for CEI and contracts already signed for ROeID. The legislation exists, eIDAS 2 being an EU regulation directly applicable following its adoption at the European level. The technology exists: Poland has delivered mObywatel, which has surpassed 12 million active users (May 2026) after a journey launched in 2017. The public will exists, 78% of Romanians want accelerated digitalisation (AtlasIntel × Edge, 2026).

What is missing: a decision-maker. Someone who can be held accountable in twelve months' time for whether the Government Digital Wallet works or not. Someone whose mandate is not shared with anyone else. Someone whose decisions are not obstructed by the mandates of two other institutions that also report to him or her.

This decision-maker, in the model proposed by Edge & Digital Nation, is the Government CIO, the second layer of digital governance. A position that, at the time of writing this document, does not exist in Romania.

And what else is missing, at least equally important: a team.

In June 2026, Edge Institute will publish "Delivering the EUDI Wallet in Romania: a governance and delivery proposal", a detailed report outlining what the team building the Digital Wallet should look like. The synthesised proposal is as follows: a dedicated implementation unit of 15–20 people in the core team, cross-institutional but with a single coordination point, modelled in accordance with the delivery unit practice academically documented since the Prime Minister's Delivery Unit in the United Kingdom (Tony Blair, 2001) and used since then in over sixty countries. The composition would be structured as follows: technical leadership (Technical Lead, Solution Architect, Security Architect), backend for the wallet and credential issuers, verification infrastructure, identification/biometrics, cryptography for WSCA/WSCD, testing and certification, legal and data protection, security and operations.

The publicly available reference budget: in Sweden, wallet development is funded through the WE BUILD consortium (~€25 million / 24 months / consortium led by Sweden and the Netherlands) plus national allocations. The exact figure for Sweden's national allocation and its Romanian equivalent will be refined in the accompanying operational report. At the national level, in Romania, the estimated order of magnitude is less than one hundredth of the total digital NRRP (National Recovery and Resilience Plan) allocation.

The report proposes an incremental delivery, starting with a functional web wallet (MVP) by 31 December 2026, which includes credentials enabling access to multiple public and private services, with a view to increasing the adoption rate. The choice is strategic because (a) it is realistically deliverable in a short timeframe, (b) web development skills are more accessible in the Romanian market than native iOS/Android development, and (c) it does not go through the opaque and slow approval process of the App Store or Google Play. The native application remains step two, after 2026.

There is no need to reinvent this model. Sweden, the Netherlands, Poland, Spain, Germany, and France already have such units, each adapted to the national context.

This is how the Implementation Agency proposed in the Edge Institute & Digital Nation Framework begins to take shape:

POLITICO Research & Analysis Division, in European Digital Capacity Index 2026 (April 2026), states explicitly: „It is unlikely that efforts to accelerate business digitalisation or AI adoption will produce durable results in the absence of functional governance structures and a sufficiently developed base of skills. The sequence implied by this index, governance first, skills second, ecosystem development last, contrasts with the way EU cohesion policy has usually approached digital development”

This is the recommended order, which explains why the AI Act, in Romania, is not the first step. The first, according to this report too, is governance.

AI is already in Romania. Pupils and students doing their homework with ChatGPT, Gemini and Claude. Teachers preparing course materials, lawyers doing case-law research, public servants summarising documents. ANAF officials, journalists, mayors, members of parliament, more and more people are using AI, at home or at work. In the iOS App Store, 3 of the top 4 downloaded apps at the time of writing this article are LLMs.

The Ipsos & Edge Institute study, „The Digitalisation of the Education System in Romania” contains updated figures on the perception of AI in education. Romanians believe that artificial intelligence brings more advantages than disadvantages to the learning process and have a predominantly positive and pragmatic perception of its use in education.

The role of AI in the learning process

The Centre for Independent Journalism runs the largest digital media literacy programme in Romania, targeting several areas of teacher training: a course accredited by the Ministry of Education for middle-school and high-school teachers, optional university courses, and teaching workshops and dissemination for students in pedagogical high schools.

„Over the past year, contact with more than 1,000 pupils, future childcare workers, educators and primary-school teachers, has shown us that uncritical adoption and use of LLMs grew from 30% in the middle of last year to almost 100% in 2026, and these behaviours are replicated among younger pupils too, even kindergarten children, who use personal assistants through voice commands. The speed of this adoption is inversely proportional to access to critical-thinking tools, which are far harder to access and scale. The latter are not commercial products; they are the efforts of individuals or institutions, especially in the non-governmental sector, not a whole-society approach, as European models show”

Cristina Lupu, director of CJI.

In this conversation, the Romanian state is a passive observer. It does not instruct: there are no published guidelines for public servants who use AI. It does not measure: there is no programme to assess the impact of AI in public services. It does not build: there is no framework that would allow ministries to pilot AI systems with measurable results. It does not protect: without a law implementing the AI Act, citizens and institutions are left exposed to potential dangers.

An international warning

On 6 May 2026, the Tony Blair Institute published an essay signed by Ott Velsberg, former Chief Data Officer of the Estonian government for nearly eight years. Velsberg formulates the clearest possible warning for states like Romania that want to skip stages:

„AI will not reform the state's operating model. It will scale its dysfunction faster than it will deliver reform.” He continues: „Many countries are still catching up on the transition from e-government to digital government, while trying to jump straight to AI. If that transition remains unfinished, AI will not compensate for it. It will expose it.”

Velsberg names six foundations that a state wishing to use AI cannot skip: legal clarity, semantic clarity, interoperability, institutional accountability, transparency, delivery capacity. Checked one by one in Romania, none is complete. „Without delivery capacity,” Velsberg writes, „transformation never leaves the pilot stage.” In the context of the government wallet with 3 teams and no command centre, and of developments in AI that the state follows more as a passive witness than as a leader and driver of action, the observation above is reinforced by the weight of an authority that built X-Road, the Estonian eID and whole-of-government data exchange.

The national context and the AI Act

The EU deadline for designating competent national authorities was 2 August 2025. By that date, only eight member states had designated their authorities on time. Romania was not among them. On 12 March 2026, seven months late, the Romanian Government adopted a memorandum designating ANCOM (the National Authority for Management and Regulation in Communications) as the national market surveillance authority and single point of contact for the AI Regulation. The enforcement model is hybrid: ANCOM as the central supervisory authority; ASF and BNR for high-risk AI systems used in the financial sector; ANSPDCP for biometrics, public order, migration, asylum, borders, justice and democratic processes; ADR as the notifying authority.

At the same time, although there is a National Strategy on Artificial Intelligence for the period 2024-2027, which is still in force today, its great shortcoming is not one of substance, but rather of form. Not being directly connected to a broad funding framework, its implementation has remained a rather abstract effort. In this context, the upcoming strategy has the opportunity to be embedded directly in the logic of the next Multiannual Financial Framework 2028-2034 and to become an instrument that produces value in public services, health and education.

On 7 May 2026, the EU Council and the European Parliament reached a provisional political agreement on the Digital Omnibus on AI. The new deadlines:

• transparency for AI-generated content, 2 December 2026;
• operational sandboxes, 2 August 2027;
• high-risk systems, 2 December 2027;
• regulated products, 2 August 2028.

The officially stated reason for the delay is that „member states were not ready”. The harmonised standards were not finalised, the competent national authorities were not designated on time, institutional capacity had not been built. Romania is the representative country for this category. The delay was set for exactly the states like Romania.

The delay is anything but an invitation to relax. It is an official acknowledgement that the lack of institutional capacity in countries like Romania is structural. Romania now has two options: to be in the group of states with an operational competent authority and functional sandboxes by the deadline, or to find itself once again among the perennial laggards.

The diagnosis, in this case, is identical to that of the Digital Wallet. The technology exists. The EU regulation exists. The Digital Omnibus amendments extend the deadlines, but do not resolve the institutional deficit. Public demand exists.

The decision-maker, however, does not exist.

The diagnosis and the urgency, reiterating, do not bear the Edge Institute's stamp alone. The "Call for Romania's Digital Transformation", launched in June 2025, was signed by six organisations with different profiles but an identical conclusion: Edge Institute, ANIS, ARMO, Concordia Employers' Confederation, Leaders in Tech, and Romanian Business Leaders. Some of them repeated the same message in the interviews for the Digital Nation report in October 2025.

The Concordia study from April 2026 on the new European Multiannual Financial Framework argues that "treating digital leadership as a single window risks creating a silo in which the enabling potential of digital technologies is isolated rather than amplified across the entire investment portfolio of the Fund". The same logic applies internally: digital governance must not be confined to a single ministry, but integrated into all government decisions that involve a technological component.

Employers' federations, entrepreneurs, civic associations, think tank members, all have called for the same thing: clear leadership, a command centre, cross-ministerial decisions. The five steps that follow reflect the operationalisation of that demand and are no longer optional.

And more importantly:

The AI framework, integrated into the digital governance model, must not be a parallel structure, but a component of the same Deputy Prime Minister for Digital Transformation mandate, which collaborates institutionally with the new State Counsellor on AI.

The only way AI governance can be done coherently is within the same apparatus that is building the digitalisation of the state.

Presidential Administration: Two Steps that Consolidate the Government's Agenda

The five points above are operational demands with public deadlines, success criteria, and costs of non-fulfilment. The agenda for the Presidential Administration, the institution with the longest time horizon in Romania's constitutional architecture, the only layer of governance capable of spanning two, three, perhaps four coalition formulas without being replaced, operates differently.

On 21 May 2026, the Presidential Administration announced the appointment of a State Counsellor with responsibilities in digital transformation, the first concrete institutional response to the agenda Edge Institute addressed to Cotroceni almost a year ago. Integrated into the coordination model proposed above, the mandate of this position becomes the bridge of continuity between Cotroceni Palace and the Deputy Prime Minister for Digital Transformation, regardless of the coalition formula that follows.

Two additional steps remain. The diplomatic network can be calibrated without a new structure. The missions in Brussels, Washington, Paris, and Tallinn are the natural points for negotiating AI and eIDAS 2 norms, for tech economic diplomacy, and for transferring expertise back to the country; an explicit mandate for existing ambassadors is sufficient.

The most important step that remains is the cross-party National Pact for Romania's Digitalisation: a mechanism that transforms the five steps above from mandate promises into national commitments, signed by parliamentary parties, academia, employers' federations, and civil society, and which survives any change of government.

Twelve months have passed since our first call and eight months since the report dedicated to governance. Two steps have been partially fulfilled by the Government. Five others remain to be taken. The economy is stagnating, inflation and the deficit remain high. Digital transformation is not seen as one of the solutions for economic growth, contrary to European evidence, but merely as a narrative for streamlining the number of civil servants in the public apparatus.

The Romanian citizen uses AI daily while the Romanian state looks on. The Government Digital Wallet has three teams in an equation with no leader. The negotiating window for the 2028–2034 Multiannual Financial Framework closes in a year and we have no digitalisation strategy capable of coherently attracting European funding.

Romania will have a new Government, and on the topic of digitalisation, as a great many voices affirm, it needs a leader and a command centre. It also needs an implementation team: a concrete one, with a single technical leader, an allocated budget, and public deadlines, to demonstrate that the model works in Romania too, not only in the rest of the EU member states. The Government Digital Wallet team, given the EU deadline and the level of operational specificity already documented in the report Edge Institute will launch, is the natural candidate to be the first such team.

Four Distinct Sources. One Single Conclusion.

These are the scores Romania brings into the new Government's mandate.

The government programme is the moment at which leadership either adopts good ideas from the past or buries them because they belonged to others. Edge Institute launches this vision document with the expectation that the digital leadership to be appointed will take ownership, including the courage to say that, without such governance, we are moving away from the European average and from the massive opportunities we can access through clear and committed decisions. We also launch this document with the promise to record, publicly, in detail, and by comparable methods, what happens if such a paradigm is not announced.

Today there are no concrete plans on governance, the Wallet, and AI. Their absence is the new Government's opportunity.