Romania and Bulgaria are EU’s AI Readiness Laggards

Posted: May 19, 2026 | In: Artificial Intelligence, Case study

A new peer-reviewed study maps AI readiness across all EU member states. The findings are blunt: Southern and Eastern Europe – and Romania in particular – face systemic barriers that go far deeper than infrastructure or investment. Here is what the research says, and what needs to change.

This should not be surprising. For years, analyses of the EU’s digital landscape have pointed to the same fault line – a persistent gap between Northern and Western European member states on one side, and Southern and Eastern ones on the other. Lower levels of digital and AI literacy, slower e-government adoption, educational disparities, weaker institutional capacity: the diagnosis is familiar.

What is new — and what should prompt genuine concern — is the specific nature of the deficit now being quantified. A 2026 peer-reviewed study published in Telematics and Informatics on ScienceDirect (“Unequal AI readiness: institutional and digital disparities in e-government across the European Union”) does not merely describe a digital divide. It maps, clusters, and names the countries that are structurally unprepared for artificial intelligence in public governance.

Romania and Bulgaria appear in the worst-performing cluster. The study calls them AI Readiness Laggards.

“Extremely low levels of digital skills, suggesting systemic barriers to AI readiness.”

What the study actually measures

Unlike broader digital economy indices, this study focuses specifically on the prerequisites for AI adoption in e-government — the machinery of public services and state institutions. The researchers identify two independent and underlying dimensions that determine whether a country is ready for AI-driven governance:

Digital Skills & E-Government Engagement – Measures citizen-side readiness: the prevalence of digital competencies in the general population and the degree to which citizens actually interact with government digitally.
Transparency & E-Government Service Availability – Measures institution-side readiness: how openly and completely public services are delivered digitally, and the transparency of government operations and data.

How the EU clusters

The study identifies six distinct clusters of member states based on their combined scores across the two dimensions. The distribution is uneven, and the gap between the best and worst performers is stark.

On one side, the AI – Ready Leaders: High digital skills. High transparency. A self-reinforcing cycle: skilled citizens demand better services; better services built more skilled citizens. Institutions here are the closest to being operationally ready for AI in governance. Here you can find countries like Denmark, Finland, The Netherlands.

On the other side, the AI Readiness Laggards: Low digital skills. Low transparency. Systemic, not surface-level, barriers. Extremely low digital skills in the general population combined with limited institutional transparency creates a compound readiness deficit. Here you can find countries like Romania or Bulgaria.

What Romania must do – and where to start

The study does not prescribe solutions, but the diagnosis makes the priorities clear. Closing the AI readiness gap requires action on multiple levels simultaneously, because the two dimensions interact: you cannot build institutional AI capacity without a population that can engage with digital public services, and you cannot drive citizen digital literacy without better, more transparent digital services to engage with.

Our view

At Zeren Software, we work at the intersection of digital transformation, software development, and institutional capacity. The study’s findings confirm what we observe in our client work across the region: the barriers to AI readiness are organisational and human before they are technical. Procuring AI tools is not the same as being ready to use them. Writing a national AI strategy is not the same as having the institutional infrastructure to implement it.

Romania can close this gap. But it requires treating digital literacy and institutional transparency not as outputs of AI adoption, but as the prerequisites for it. The study makes clear that the countries which are genuinely ready started building those prerequisites long before AI became the policy priority it is today.

The readiness gap is a shared problem. Closing it requires pressure and example also from the private sector, not just government initiative.

A Data Architect Close to Data Swamps and Mountain Trails

Posted: May 12, 2026 | In: Data Engineering

What does a Data Architect actually do all day? For Sânziana, the answer starts somewhere between a data swamp and a mountain trail. Actually, this contrast tells you more about her than any job title could. Sânziana has spent her career moving between banking, retail, and now international manufacturing – always at the intersection of raw data and business sense. Today she works as one of Zeren’s senior consultants, leading a Business Intelligence engagement for a large manufacturing client in the Nordics.

From tool to language

Her relationship with data started at university. It has never really stopped since then. “Every job, every freelance project, revolved around data,” she says. “How we load it, how we transform it, how we use it to produce reports that are genuinely useful to the business.”

But over time, something shifted. She moved from the mechanics of loading and structuring data to something harder to describe.

“I moved from simply loading data to understanding what it could do for people. That shift – from technician to translator – is what I enjoy most.”

The magic that isn’t magic

Ask someone outside IT how data gets into a dashboard. Fost probably, the honest answer is usually: “I assumed it just… appears.” Sânziana has heard this many times.

“The most challenging part is helping them understand it’s a multi-step process,” she explains. Sources that seem completely unrelated – CRM data, supply chain systems, financial records – all have to be pulled together, profiled, cleaned, and reshaped before a single chart renders. Load everything without care, and you end up in what the industry calls a data swamp.

“It may look like magic. But it’s really the combined work of many people, each with their own specialism, working across multiple layers before anything reaches the business.”

On Zeren

Eventually, Sânziana came to Zeren looking for something her network couldn’t give her: access to projects at a different scale.

“Zeren opened my horizon toward international projects – more complex engagements, for clients I wouldn’t or couldn’t have reached on my own.”

In fact, if she had to explain Zeren to a fellow data architect in plain terms:

“For the freelancer who wants to grow beyond their existing network, Zeren opens doors. To interesting clients, new industries, projects you’d never have found alone.”

Off the clock

Lastly, away from the data warehouse, Sânziana follows advice from an unlikely source – her son. “I go outside to touch grass, as he says. Look at the sky, go for a hike. I love the mountains. That’s my antistress ritual — nature and the mountains.”

Indeed, the people building the infrastructure modern business runs on are also the people with many hills to climb and skies to look at.