Sari la:
This week, at its annual summit, ANIS (the Employers’ Association of the Software and Services Industry) published a comprehensive diagnosis of Romania’s domestic ICT industry. Produced by Implement Consulting Group, the report illustrates a reality that most of us sensed but few could articulate clearly: Romania is closer to its potential than it believes, yet more exposed to missing it than it cares to admit.
Under the heading of “strategic planning,” Romania is still struggling. Comparative analyses, set against a regional backdrop, targeting industries of strategic importance to Romania and exploring possible trajectories for the country to assert itself in this equation, are rarely found. ANIS offers precisely such a document, and the public policy visions reflected in the study must become the subject of broad societal debate.
What follows is an overview of the material, along with several important ideas we extracted “between the lines” of its 60 pages of data, comparative illustrations, and interviews with industry leaders. We start from a question that is as simple as it is complex: what exactly is holding us back, and what could propel us forward?
The Starting Point: A Mature Industry at a Crossroads
Despite the obstacles, Romania’s IT industry has consolidated at a remarkable pace over the past two decades. Today, the ICT sector contributes approximately 8% of Romania’s total economic value added, one of the highest shares in the European Union. Apparent labour productivity in IT exceeds the average across all sectors by 16%. Between 2021 and 2023 alone, the sector grew by 17% annually. Even in 2024, a year of significant macroeconomic pressure, growth reached 7%, above most other sectors.
On paper, this industry is doing well. The model that enabled this growth, however, is unsustainable. Outsourcing, multinational service centres, cheap labour and low costs; all these premises create a growth paradigm that has already reached its limits. Salaries are rising. In the meantime, regional competition is intensifying. International clients are increasingly less interested in low-cost execution capacity and increasingly more interested in partners with their own products, validated intellectual property, and solutions that solve a concrete problem.
The report puts its finger precisely on this wound: without an innovation paradigm to mobilise the industry, it risks falling into a deadlock unable to keep pace with increasingly pronounced global dynamics. At the same time, in the context of the new “Industrial Revolution” we are traversing, a historic window of opportunity allows Romania to make a major leap in the regional equation, and beyond.
What stands between Romania and that leap? The report identifies four concrete barriers.
A “Testing Culture” That Remains on the Periphery
Among Central and Eastern European countries, Romania registers yet another negative record: a fragile ecosystem for product testing and product innovation. On the performance of SMEs in product innovation, indexed against the European average, Slovenia records a rate of 159%. Estonia stands at 126%, the Czech Republic at 107%, and Poland just under 30%. Romania? A mere 12.3%.
At less than one-eighth of the European average, Romania ranks among the countries that create the fewest ecosystems and infrastructures for SMEs and start-ups to test, validate innovative solutions and bring them to market through this mechanism. As the ANIS report reflects, there is a major access gap for SMEs to sophisticated AI platforms, IoT networks, 5G environments, and digital testbeds. Digital inequalities are thus also visible in this segment.
Real-world testing costs money. And for a small company with no safety net and no access to public data, the risk of a failed pilot is prohibitive. Estonia, for example, addressed such challenges through programmes like the Digital Testbed Framework, built in 2021 on the existing e-government infrastructure, with near-zero capital costs. Pilots immediately reach hundreds of thousands of citizens, companies gain reference and credibility, and the state receives solutions for free.
The Notion of “Ecosystem” Is a Rare Occurrence
In Romania, the fragmentation of digital governance translates into the market and the ecosystem as well. As the study reveals, Romania is the country in the region with the lowest rate of partnerships that its companies leverage to develop innovative projects. At 16%, Romania sits 11 percentage points below the EU average. Meanwhile, Estonia stands at 39%, and Hungary and the Czech Republic reach 34%.
What does this mean in practice? It means that Romanian start-ups with a good product do not reach their first corporate client, not because the product is weak, but because the mechanisms that would facilitate the meeting between a “pioneer” and an organisation seeking a solution from them simply do not exist. Without a first client, there is no reference. Without a reference, there is no investment. Without investment, there is no scaling. The circle closes before it begins. The absence of a primary customer thus becomes the most frequently cited barrier by industry leaders — even ahead of talent shortages.
Research and Development: The Cinderella of Government Programmes
As a percentage of GDP, Romania invests less than one percentage point of its annual consolidated budget in research and development. In practice, approximately 0.3% of GDP, the lowest share in the European Union in this sector. Despite holding strategic assets, including a substantial tech diaspora connected to the world’s major ICT networks, Romania fails to leverage them in any meaningful way. A second problem compounds this equation: the absence of a solid governance architecture in the research and development sector.
While Romanian companies make a sustained effort to innovate, their initiatives do not register as R&D activity, also because intellectual property rules are not coherently defined and tax incentives are patchy. The current facilities are largely inaccessible in practice, owing to their complexity and lack of clarity, as well as to prohibitive bureaucracy. As a direct consequence, the innovative output of Romanian IT teams ends up being registered and monetised in countries like Ireland, the Netherlands, and Luxembourg, countries that offer a predictable tax framework. While value is created in Romania, the fiscal benefit remains elsewhere.
The Romanian State Buys Volume, Not Innovation
Romania allocates 5% of public procurement to innovation, compared to the European average of approximately 9% and Estonia’s rate of 14%. Behind these static figures lies a strategically bankrupt (in)decision: the state, the largest buyer in the economy, does not purchase domestic solutions on the market. The legal mechanism that would change this equation, the Innovation Partnership stipulated in Law no. 98/2016 on public procurement, has never been used in sectors such as AI or cloud.
The Opportunities: Why the Window Is Now, Not in Three Years
The barriers are plain to see. So are the opportunities. And seizing them is becoming a matter of national urgency. As Robert Berza, Executive Director of Edge Institute, noted in previously published materials on governmental AI, Romania needs to sharpen its compass for opportunity in the sphere of digital transformation and new technologies, and make an ambitious leap.
The ANIS report identifies, in a valid manner, four mega-trends that create this landscape of opportunity: AI, cloud, cybersecurity, and the Internet of Things. Across all four sectors, there is potential, both in human and technical terms, for Romania to become a regional entrepreneur, a European innovation hub, rather than merely a provider of execution capacity. Across all these sectors, there are conditions for Romania’s business ecosystem to produce, under intellectual property ownership, authentic solutions with amplified added value.
This is by no means an optimistic assessment. The data confirm this reality. Romania has an industry of approximately 220,000 IT specialists, an industry that has sustained consistent growth over the past decade. This industry has also created an ecosystem of entrepreneurial companies that have passed the test, becoming suppliers of high-quality products to major clients in Western Europe, the US, and beyond. From this we can deduce that the competence infrastructure exists. What is missing is the ecosystem infrastructure: the primary customer, early-stage capital, the fiscal framework, and the testing environment.
In this regard, the report presents two important comparative analyses. If Romania were to reach, proportionally, the size of Poland’s IT industry, the impact would amount to an additional approximately €38.8 billion in GDP, 294,500 jobs, and €88 billion in cumulative tax revenues. In the Czech scenario, Romania would register an addition of approximately €5.9 billion to GDP, 44,500 indirect jobs, and €13.3 billion in cumulative tax revenues. It is worth noting that these figures capture only the direct impact of the IT industry on the economy, not the spill-over effects that such a development would generate towards other sectors.
One detail from the report deserves separate attention. Romania has approximately 220,000 IT specialists. Czechia has 235,000, with a population almost two times smaller. With roughly the same talent pool, the Czechia produces ICT output that is 19% larger. The difference comes from what these specialists do: own products versus execution capacity, IP registered locally versus value created here and taxed elsewhere. This is, at bottom, the necessary transition the entire report talks about.
Read on a per-capita basis, the figures say something further: the gap relative to Czechia is the one that demands urgency, because it is not about volume but about structure, and it requires substantive interventions.
Moreover, the raw number of specialists conceals another problem. A product-oriented industry needs not only people who code, but also people who sell, position, and scale. Romania does not train or retain enough people who know what to do with a product once it has been built. Czechia has partly resolved this through its accelerator ecosystem and through domestic corporate demand, which has produced a generation of product managers and tech entrepreneurs with real commercial experience. Romania has not yet substantially developed that domestic demand, another point underlined in the report. This is precisely what the state could do as a primary customer: buy more Romanian technology (5% versus the EU average of 9%) and build a stronger ecosystem of teams that know how to develop, validate, and sell it.
AI: The Catalyst the Report Sees But Cannot Fully Measure
In the absence of a robust foresight and analysis framework for the impact of artificial intelligence, the report provides solid working premises. According to the study, the widespread adoption of AI could contribute approximately €14–16 billion annually to Romania’s GDP, equivalent to 5% of its total. Naturally, to realise such a contribution, a solid national AI policy is required, and a ten-year horizon of progressive adoption becomes vital.
The magnitude of these figures is driven by three key mechanisms: productivity gains, efficiency through the reduction of repetitive administrative tasks, and the tactical reallocation of the time saved. According to the study, roughly 1 in 2 jobs in Romania will, in the medium and long term, involve some form of structured interaction with AI tools. Full replacement by artificial intelligence is projected to affect a segment of around 4% of the current workforce.
As Alexandru Dănescu and Roland Kristo elaborated in a report published for Edge Institute, automation in the labour market requires a robust policy to build the capacity of companies, individuals, and the economy. In practical terms, the difference between a company that adopts AI in 2026 and one that adopts it in 2028 is not merely a two-year gap but a generational competitive advantage.
There is, however, an underlying layer that the report, finalised in December 2025, could not fully reflect, though it managed to hint at it between the lines: in the meantime, AI has become not only an opportunity for the domestic ICT industry but also an accelerator of the very transition the report discusses. In practical terms, a proof of concept that in 2022 took six months and cost approximately €50,000 can now be built in a few weeks for a few thousand euros. Market research that once required a team and months of work can today be completed, with the help of a well-prompted agent, in just a few hours.
Prototyping is thus becoming a far more accessible process, though the market is already feeling the full force of these developments.
So, What Is to Be Done?
A simple calculation in the report deserves particular attention. According to the report, if Romania maintains its historical annual growth rate of 17% and eliminates the bottlenecks that reduced growth to 7% in 2024, the country can reach the scale of Poland’s IT industry within approximately six years. A simple mathematical exercise illustrates this reality. That said, Romania’s roadmap remains vast. First and foremost, however, three constants must be preserved: talent remains available at competitive salaries, growth capital does not contract, and market demand continues to expand.
The instruments proposed in the report are precisely those that sustain these three conditions. Without them, the baseline scenario is stagnation, an industry growing at 2% annually, in line with the rest of the economy, missing the AI wave entirely. The window of opportunity that AI represents does not stay open indefinitely. Any public policy projection must, in practice, start from this consideration. Romania must therefore adopt several instruments that Edge Institute has discussed at length in various publications.
First, developing a national AI sandbox, a regulated environment for testing and co-creation between private sector actors, public institutions, civil society, and academia, is an imperative. This format could subsequently be replicated at the regional level to help reduce regional disparities. Likewise, launching innovation pilots and testbeds within clusters adapted to regional economic realities and profiles represents an important opportunity. This national AI sandbox is also a legal obligation under the AI Act, with a deadline of August 2026, one that Romania has not yet begun to address.
Second, Romania must deploy creative measures to stimulate its ICT talent. The idea of a “Digital Civic Service” as a dedicated initiative for experience exchange in Romania aimed at the “tech diaspora” could serve as such a platform. Alexandru Dănescu and Roland Kristo also discuss the need for a National AI Observatory, one that must, of course, also cover the ICT sector. Additionally, the creation of learning, upskilling, and innovation hubs, modelled on France’s Campus Cyber, would be a worthwhile initiative.
Third, a solid governance architecture is essential. From establishing a dedicated Chief Information Officer office for the government, under the authority of the Deputy Prime Minister responsible for digital transformation, to creating a foresight unit at the centre of government, such measures are indispensable. Digital economic diplomacy is equally important: building a network of “digital attachés” and special tech envoys to advance the interests of the local ICT industry abroad is a necessary step.
The ANIS and Implement report says that Romania has a solid IT industry standing at a crossroads. That the transition towards innovation is possible. That the instruments to facilitate it are known and adaptable. That the barriers are not insurmountable.
As we emphasise on every occasion at Edge Institute, the window is open; the diagnosis is on the table. The instruments are documented. The window of opportunity is open, but not indefinitely.
What remains “pending” is the decision-making capital.