Moldova’s IT industry has just crossed $1 billion in turnover. At the same time, many engineers feel the market is becoming harder. AI is part of the reason, although probably not in the way most people think.
Inside the industry, reactions to AI range from excitement to quiet anxiety. If you are curious how AI adoption looks inside a fast-growing outsourcing destination, the Moldovan experience offers an interesting perspective.
A Growing Industry, Not a Declining One
First, a bit about how the IT landscape looks like today. Far from shrinking, Moldova’s IT industry is expanding. The Moldova Innovation Technology Park (MITP), the backbone of the country’s tech ecosystem, reached a historic milestone in 2025: $1 billion turnover · 25,803 employees (+7% YoY) · 2,725 resident companies (+27% YoY) · ~88.5% of revenue from exports.
Over a longer horizon, the numbers are even more striking: revenue has grown ~10x since 2018, employment has grown ~4x in the same period.
This is not what disruption looks like. At least not yet. The Moldovan IT sector continues to expand despite, and partly because of, the AI wave.
Why the Market Feels Worse Despite Industry Growth
Despite the growing number of companies and increasing revenues reported by MITP, the perception among many engineers in Moldova is noticeably less optimistic. Many feel that there are fewer attractive projects available, and that finding a company, role, or project that truly matches their expectations has become significantly more difficult. Compared to a few years ago, engineers feel they have less choice, less flexibility, and less negotiating power.
This can be explained through a combination of factors.
During the pandemic and immediate post-pandemic years, the global IT outsourcing market experienced an exceptional growth cycle. Demand for engineers was significantly higher than the market could realistically supply.
In Moldova, strong engineers often had the luxury of choosing between multiple projects simultaneously: preferred technologies, preferred teams, remote flexibility, compensation packages, and working conditions.
At one point, a good engineer could realistically choose between several competing offers at the same time.
Companies were willing to overpay simply to secure delivery capacity.
While this created a very favorable market for engineers, it also distorted expectations on both sides. Salaries accelerated aggressively, while businesses accepted costs that later became difficult to justify economically.
In hindsight, many players in the industry started treating an exceptional market cycle as if it were the new normal.
During the last couple of years, the situation has changed visibly. The number of projects dropped, and companies became significantly more cautious with spending. Part of this comes from the global economy, uncertainty, and businesses becoming more focused on efficiency instead of aggressive growth. And of course, AI is now part of this equation as well.
While AI has existed for some time already, its real influence on the market has become noticeably stronger during the last 6 months.
Teams are increasingly expected to deliver more with fewer people. The pressure is felt especially around junior and repetitive work.
At the same time, companies are becoming much more careful about what kind of engineers they actually need.
Still, despite all this, the market continues to grow. That is the interesting part. According to the MITP report, employment is still growing (+7%), but growth is slowing compared to previous years. Existing companies are not the main source of new jobs anymore; growth comes mostly from new entrants. At the same time, revenue per employee increased by ~21.7% year over year. This combination is critical: Higher productivity + slower hiring growth = early AI effect.
Why AI Hasn’t Replaced Engineers (Yet)
The intuition behind AI-driven disruption is simple: If AI can write code, fewer developers are needed. But this logic breaks down in real-world environments. As highlighted in multiple situations, AI works best in “clean,” greenfield systems. In reality, many companies operate in:
- Legacysystems; · poorly documented architectures; · multi-system integrations; · strict uptime requirements.
Moldova’s IT sector is heavily exposed to exactly these kinds of environments, especially in outsourcing and enterprise projects. This creates a paradox: AI increases expectations ofproductivity but does not eliminate the need for engineers. Instead, it sometimes increases demand for: integration of work, system understanding, architecture, and governance.
This aligns strongly with the findings from the 2026 engineering report: AI is reducing pure implementation work, but increasing demand for problem-solving, architecture, and governance. In other words: less typing, more thinking.
The Moldova-Specific Advantage: Lag Is a Feature, Not a Bug
Moldova has a structural characteristic that slows disruption: Adoption lag. And in the context of AI, this is not necessarily negative.
From both research and practical observation: · Large organizations have a “braking distance” when adopting AI; · Regulated and industrial sectors adopt even slower; · Integration complexity delays full automation.
This creates a window of stability. In Moldova, where a significant share of IT work is tied to outsourcing, enterprise systems, industrial and near-hardware domains …the pace of disruption is naturally moderated.
ISD’s Perspective of the Market
The picture is more nuanced than the public narrative often suggests.
So far, there is no consistent evidence of AI broadly replacing software engineering teams. There are anecdotal examples of reductions among some companies in Moldova, but nothing that looks systemic.
What is becoming visible instead is a gradual change in expectations, delivery models, and the type of engineering work companies value the most:
- Demand Is Still There, But Expectations Changed.
- Clients increasingly expect lower software development costs because of AI, and in many situations these expectations are unrealistic. AI is not eliminating demand. It is reshaping pricing pressure.
Early-Stage Development Is Being Disrupted.
Startups increasingly build MVPs with “vibe coding”, less demand for full dev teams at the earliest stage. This is a real shift. But it affects: who hire developers, not necessarily how many are needed overall. - Sector Matters More Than Ever
ISD operates significantly in industrial and automation domains: slower AI adoption, hardware dependencies, and higher complexity.
Moldova’s competitive advantage is more likely to come from quality than quantity.
In practice, many businesses are not simply looking for somebody who can generate code faster. They are looking for partners who can understand the business problem; navigate complexity; and deliver outcomes responsibly
What Becomes Moldova’s Competitive Advantage in an AI World?
Despite having one of the highest ICT workforce densities in emerging Europe, Moldova remains a very small country. Around 4.4% of the country’s workforce is employed in the ICT sector, the second-highest ratio in emerging Europe after Estonia. Yet Moldova simply cannot compete globally on volume alone. Countries like India or larger Eastern European markets operate at a completely different scale. Moldova’s competitive advantage is more likely to come from quality than quantity.
In practice, many businesses are not simply looking for somebody who can generate code faster. They are looking for partners who can understand the business problem; · navigate complexity; · and deliver outcomes responsibly.
In an AI-driven world, trust, adaptability, and systems thinking may become more valuable than raw coding speed.
This may become one of Moldova’s biggest advantages.
The result is a less immediate impact from AI compared to pure software environments. This reinforces a broader point: AI disruption is uneven and far from universal.
Conclusion: No Disruption (Yet). But No Illusions Either
Moldova’s IT industry today looks resilient: record revenues, growing employment, strong export engine. AI has not broken the model, but it is quietly rewiring it.
The real risk is not that AI will suddenly eliminate jobs. The real risk is slower and more subtle: companies and engineers who fail to adapt will become irrelevant, even if the industry keeps growing. That is a much harder problem to detect and a much harder one to solve.
