A Prize on the Table, a Region Behind the Curve
The number that frames Central Europe’s AI moment comes from McKinsey’s June 2026 analysis of the region’s AI potential, and it is large enough to warrant serious attention from every boardroom in Warsaw, Prague, Budapest, and Bratislava. AI could help uncover between two hundred and eighty billion and seven hundred billion euros in economic value across the region, equivalent to six to fifteen percent of the region’s total net turnover. Framed differently, the gap between Central Europe capturing that value and Central Europe failing to capture it is the difference between a regional economy that competes effectively in the next decade of global industrial competition and one that watches that competition from a progressively weaker position.
That framing is not dramatic for the sake of it. It reflects a genuine structural juncture. Two realities sharpen the urgency that McKinsey’s Balázs Czímer and Lieven Van der Veken document in the analysis. First, the global pattern of widespread deployment without widespread impact is already visible in the region: eighty-eight percent of companies globally have deployed AI in at least one function, but ninety-four percent have not achieved a significant impact on earnings before interest and taxes. A long list of pilots may signal that companies are making progress, but these small efforts rarely change end-to-end performance at the level that moves an income statement.
Second, Central Europe trails Western Europe in enterprise AI adoption by sixteen percent, and approximately sixty percent of Central Europe’s economy is tied to sectors where scaling AI is most difficult, manufacturing, engineering and construction, consumer goods, and retail, precisely the domains where operational complexity runs deepest and digital maturity has historically lagged. Both observations together describe a region that is not absent from the AI conversation but is not yet converting its presence in that conversation into the economic outcomes that the conversation is ultimately about. Understanding why the gap exists, what it would actually take to close it, and which companies and countries in the region are already demonstrating the path forward, is the analytical task that the McKinsey analysis advances and that this long read examines with the specific context of Central Europe’s industrial structure, talent dynamics, and policy environment in mind.
The Pace of AI Advancement Has Already Outrun Business Planning
The first analytical anchor the McKinsey report places is one that deserves more attention than it typically receives in regional AI discussions, which tend to focus on adoption rates and use case inventories rather than on the pace of the underlying technology itself. Between 2019 and 2022, leading large language models from OpenAI, Google, and Anthropic advanced at roughly two index points per year on the Artificial Analysis Intelligence Index, a composite benchmark spanning reasoning, knowledge, mathematics, and programming. From 2024 onward, gains exceeded twenty points per year and kept accelerating, representing a tenfold increase in the pace of advancement since the introduction of ChatGPT in late 2022. Overall AI capability is now doubling approximately every twelve months.
The consequence for business planning is structural rather than merely statistical. Strategy, capital allocation, and operating model redesigns are usually planned over multiple years. When core technology improves materially within a single year, the assumptions embedded in those plans lose relevance faster than the plans can be revised through normal governance cycles. A Central European manufacturing company that conducted an AI readiness assessment in 2023 and concluded it would return to the topic in its next strategic cycle in 2025 arrived at that cycle to find that the technology it was planning around had advanced far beyond what the 2023 assessment assumed was possible, while competitors who had been experimenting continuously throughout that period had accumulated learning, proprietary data, and organisational capability that the non-experimenting company had not.
The Stanford Human-Centered AI Institute’s 2025 Artificial Intelligence Index report documented the specific capability expansion that has occurred across this period: as recently as 2019, AI was effective only at narrow, well-defined tasks such as image classification, basic language processing, translation, and speech recognition, and it fell short in multitask understanding, advanced mathematics, and cross-modal reasoning. Within six years those boundaries dissolved. Tasks that previously required specialised human expertise, including routine analysis, document processing, and structured decision-making, are now within the reach of AI systems, shrinking the set of activities in which human involvement is essential. For Central European companies still treating AI as a narrow efficiency tool for specific defined tasks, the speed of this capability expansion means that their mental model of what AI can do may already be outdated by one to two generations of model advancement.
Why Waiting Compounds the Disadvantage
The McKinsey analysis makes a specific and important argument about the relationship between adoption timing and competitive position that is easy to overlook amid the quantitative framing of the opportunity. Early movers in AI adoption learn where models fail and where they can be trusted, build proprietary data flows that improve model performance for their specific contexts, redesign decisions around human-machine interactions that produce better outcomes than either alone, and develop institutional routines for rapidly adjusting workflows as the technology continues to advance. These are not one-time advantages that latecomers can simply acquire by purchasing the same tools at a later date. They are compounding organisational capabilities that accumulate over time and that become harder to replicate the longer they have been developing in competitors who moved earlier.
Delaying adoption for the sake of optionality, waiting until the technology matures further or until best practices are more clearly established across the industry, sets companies back not only because competitors will have advanced in the interim but because the organisation itself will have fallen further behind in the learning curve that AI deployment requires. The learning, in other words, is not separable from the doing. A company that waits three years to begin serious AI deployment does not simply start from where early movers started three years earlier. It starts from a position where the technology it is adopting is more advanced and more complex, where the organisational capability required to use it effectively has not yet been developed, and where competitors who have been building that capability continuously are three years further along in their own compounding advantage.
For Central European companies in manufacturing, engineering, and retail, the specific risk the McKinsey analysis identifies is not sudden collapse of the kind that digital-native competitors can inflict on purely digital businesses, where a new entrant can replicate a product and undercut an incumbent’s pricing within months. The risk is slow erosion: a gradually widening gap in operational efficiency, product quality, customer responsiveness, and cost competitiveness that becomes apparent only when it is too large to close through the normal pace of internal transformation programmes. The Duolingo example the McKinsey authors cite, in which the company lost most of its market value after investors concluded AI-native alternatives could replicate its offering faster and cheaper, illustrates the endpoint of that erosion when it reaches a tipping point. Central European manufacturers are unlikely to face a Duolingo-scale collapse from AI disruption. They are at risk of the quieter, more insidious version: being outcompeted margin point by margin point, contract by contract, over years, by rivals who embedded AI into their operations while others were still deciding whether to run a pilot.
The Structural Challenge: Industrial Strength in the Hardest-to-Scale Sectors
The specific economic geography of Central Europe’s AI opportunity creates a challenge that is different in character from the AI transformation challenge facing Western European economies with larger technology, financial services, and professional services sectors. McKinsey’s analysis found that the largest potential gains in the region are in advanced manufacturing and consumer goods and retail, sectors that together represent the backbone of Central Europe’s productive capacity. These are not sectors where AI can be layered onto digital channels that already exist and scaled from there. They are sectors where, as the analysis notes, much of the value of AI will come from embedding AI into physical operations rather than layering it onto existing digital infrastructure.
In manufacturing, AI can optimise production planning to improve factory utilisation, refine input material management to reduce yield loss, and improve maintenance scheduling to reduce unplanned downtime. These are genuine, measurable, financially material value levers. They are also operationally complex to implement. Manufacturing floors generate data from a heterogeneous collection of sensors, machines, enterprise resource planning systems, and quality management platforms that were not designed to feed AI systems and that require substantial data engineering work before any AI model can be productively trained on them. The OECD’s 2026 analysis of AI uptake in EU manufacturing found that AI adoption in the sector remains modest and highly fragmented, with adoption skewed toward discrete, well-defined use cases such as predictive maintenance, quality assurance, and supply chain optimisation, while broader, end-to-end operational AI integration remains rare even among the most digitally advanced manufacturing companies in the region.
The complication is that Central Europe’s dominant manufacturing base is disproportionately oriented toward the most operationally complex manufacturing categories. The automotive supply chain, which employs hundreds of thousands of workers in Poland, the Czech Republic, Slovakia, and Hungary, involves just-in-time production systems, multi-tier supplier networks, and quality standards that create genuine complexity for AI deployment at a scale that simpler manufacturing contexts do not encounter. The engineering and construction sector, another large contributor to Central Europe’s net turnover, operates on project-based rather than continuous production cycles that make the data collection and model training infrastructure of AI deployment harder to sustain across the gaps between projects. The McKinsey analysis found that approximately sixty percent of Central Europe’s net turnover is in industries where just seventeen to eighteen percent of AI adoption has reached the scaling phase, a figure that captures the extent to which the region’s economic strength is concentrated in the sectors that have proven hardest to move through the AI adoption curve globally.
The Adoption Gap That Tools Cannot Explain Away
One of the more instructive observations in the McKinsey analysis concerns the gap between AI tool adoption and AI-driven change. In software development, one of the most digitally advanced fields in any economy, approximately ninety percent of developers now use AI coding tools, but only twenty to thirty percent of that total have changed how they work as a result. The overall productivity improvement from AI coding assistance has consequently been less than fifteen percent, not because the tools are ineffective but because tool adoption and workflow redesign are not the same thing, and the productivity benefit of AI derives from the latter rather than the former.
This distinction applies with even greater force in Central Europe’s dominant sectors. A manufacturing company where engineers use AI to draft maintenance reports faster has adopted AI. A manufacturing company where the maintenance planning process has been redesigned around AI-predicted failure probabilities, with the human engineering team now focusing on judgment, exception handling, and process improvement rather than routine data collection and report generation, has changed how work gets done. The first company will report AI adoption in any survey. The second company will report AI impact in its financial results. The gap between the two is not primarily a technology gap. It is an organisational design gap, a willingness to redesign processes rather than simply augment them with new tools.
The CEPR analysis of AI adoption across European firms found that digital maturity, specifically prior investment in cloud computing and robotics, is the strongest predictor of effective AI adoption, precisely because these prior investments reflect an organisational capability for technology-driven process change rather than simply technology acquisition. Central European firms in manufacturing and construction that have invested in automation, digital production management systems, and cloud-based enterprise platforms are substantially better positioned to move from AI tool adoption to AI-driven operational redesign than firms that have not built that foundational digital infrastructure. The implication is that the AI adoption gap between Central and Western Europe is not solely a function of AI strategy differences. It is partly a function of the accumulated digital investment gap that preceded AI, and closing the AI gap requires addressing both layers simultaneously.
From Pilots to Profit: The Three Moves That Separate Leaders
McKinsey’s analysis identifies three distinct moves that distinguish AI leaders from AI experimenters in Central Europe, and the first and most foundational is the hardest to operationalise: setting the right strategic scope before any AI deployment begins. The leaders do not start with technology and work backward to business impact. They start with an enterprise-wide assessment of where AI can redesign end-to-end processes, sharpen decision quality, and materially move the needle on growth, cost, risk, or customer experience, and they identify a small number of transformative bets from that assessment rather than a large portfolio of exploratory pilots.
The contrast this approach creates with the more common pilot-accumulation approach is not merely aesthetic. When organisations deploy AI in fragmented ways that do not support end-to-end economics, the individual pilots rarely create measurable financial impact regardless of their technical quality. McKinsey’s case study of a Central European bank that set the ambition to become the region’s first fully agentic bank illustrates the alternative. A diagnostic process identified fourteen core domains where performance could be significantly improved with AI, representing potential revenue growth of thirty to forty percent and cost optimisation of fifteen to thirty percent. More than fifty management sessions translated those findings into a prioritised transformation roadmap based on potential impact. The investment was not primarily in technology, though technology investment followed. It was in the strategic clarity to commit to a small number of high-value domains and redesign them comprehensively rather than spreading AI experimentation thinly across the organisation.
The CEE AI Index 2026, published by Tech.eu in June 2026, found that governance has become a critical differentiator across Central and Eastern European markets: while most countries in the region have introduced national AI strategies, only a smaller group has developed the institutional capacity required to implement them. Estonia emerged as the region’s most institutionally mature AI ecosystem, combining advanced digital public services, strong enterprise adoption, and a high concentration of AI talent. Poland remains the region’s largest AI market, leading in research output and workforce scale. The conclusion the report draws, that competitive advantages are distributed across research, talent development, infrastructure, governance, and market scale, with no single country leading across all categories, applies equally at the enterprise level: the companies that will capture Central Europe’s AI prize are not those with the most advanced AI tools but those with the governance capability to direct those tools toward the highest-value domains consistently over time.
Building Solutions That Live Inside Workflows, Not Beside Them
The second move McKinsey identifies is the transition from AI as a tool that employees use optionally to AI embedded directly into core workflows with clear accountability for measurable outcomes. This is the phase where most Central European organisations stall, and the specific failure mode is consistent enough across industries and geographies that it warrants naming precisely: promising ideas remain unrealised because no one owns the transition from concept to working product, and because the technology, business, and risk functions involved have not agreed on who is empowered to make the decisions needed to move forward.
The McKinsey case study of a Central European insurer that deployed hyperpersonalised campaigns across more than three hundred microsegments illustrates the scale of impact that this transition can deliver when it is managed well. An AI knowledge assistant scanned more than one thousand policy documents, voice-recognition agents provided sales coaching to frontline staff, and ninety-five percent of calls were automatically reviewed for quality and compliance. Reach rates improved three to four times, conversion rates improved two to three times, and call and processing times fell by twenty-five percent. None of these results came from running a pilot alongside the existing workflow. They came from redesigning the workflow itself around AI capabilities, with the legacy manual process replaced rather than augmented.
The agentic enterprise model that McKinsey describes as the next practical step for organisations that have completed this workflow redesign reflects where the technology is heading. Rather than operating as standalone assistants, AI systems are moving into human-supervised workflows as specialised agents. In banking, groups of agents collaborate in squads that handle document ingestion, credit memo generation, compliance validation, and client communications, with human credit managers and workflow specialists overseeing the process and intervening where judgment is required. The operational implication for Central European companies is that the build phase is not a project management exercise but a product development one, requiring minimum viable product definitions, explicit success metrics, real user involvement from day one, and rapid iteration to validate impact rather than the slow, committee-driven approval cycles that characterise traditional IT project delivery in the region.
Redesigning the Enterprise, Not Just the Function
The third move, and the one that distinguishes companies that achieve sustained AI-driven performance improvement from those that plateau after an initial wave of successful use cases, is the deliberate redesign of the enterprise itself around AI at scale. This is not primarily a technology challenge. The McKinsey analysis is explicit that the barriers at this stage are structural rather than technical: many organisations attempt to deploy advanced models while keeping legacy governance structures, siloed data sets, fragmented technology stacks, and traditional role definitions that make it impossible for AI to deliver end-to-end rather than task-level impact.
The Aviva case study that McKinsey references, in which the UK insurer transformed its end-to-end claims journey by hiring more than fifty data scientists and engineers, deploying more than eighty machine learning models across damage assessment, fraud detection, and repair routing, and investing in more than forty thousand hours of training for existing staff, illustrates the scale of organisational change that genuine AI-driven transformation requires. Assessment times fell by twenty-three days, complaints fell by sixty-five percent, and the customer satisfaction score improved by a factor of seven. These outcomes were not produced by deploying AI tools alongside an unchanged claims process. They came from using AI deployment as the occasion to fundamentally redesign how claims are assessed, routed, and resolved, with the human workforce’s role reshaped around the specific judgment, exception handling, and customer interaction capabilities that AI systems cannot yet replicate.
The Central European software company case study in the same analysis, where an agentic approach to code modernisation freed up twenty to thirty percent of developers’ time and improved EBITDA by thirty to forty percent through a train-the-trainer programme reaching more than fourteen hundred employees, illustrates that this scale of organisational redesign is achievable in the region with the right governance, capability-building programme, and leadership commitment. McKinsey’s Rewired framework identifies the six enablers of sustained AI value: strategy, talent, operating model, technology, data, and change management. These form an integrated system in which weakness in any single area constrains all the others. Central European organisations that invest in technology without addressing talent and change management will find their technology assets underutilised. Those that invest in change management without addressing the data foundations needed for AI models to perform reliably will find their organisational readiness wasted on tools that underperform in production.
The Infrastructure Being Built Beneath the Opportunity
While enterprise-level AI adoption strategy is the primary focus of the McKinsey analysis, the enabling infrastructure being built across Central Europe through both public and private investment is changing the conditions under which that strategy can be executed. The EuroHPC Joint Undertaking’s selection of AI Factories in the Czech Republic, Poland, and Romania in October 2025, alongside AI Factory Antennas in Hungary, Slovakia, and Latvia, represents the most significant public investment in Central European AI compute infrastructure since the EU’s digital agenda took its current form. The Czech AI Factory, built on the KarolAIna supercomputer at IT4Innovations National Supercomputing Center, is specifically designed to serve Czech and European AI industrial, academic, and public institutions, giving the region’s AI startups, SMEs, and researchers access to AI-optimised high-performance computing that was previously available only through foreign cloud providers or at major Western European research institutions.
The Czech Republic has gone further, with its government approving the AI Gigafactory CZ project in November 2025, a project with a budget exceeding ninety billion Czech crowns designed to create central European AI infrastructure at a scale comparable to the Gigafactory facilities being developed in Western Europe through the Commission’s InvestAI mechanism. The Ministry of Industry and Trade began working on the final application to EuroHPC in spring 2026 while simultaneously initiating negotiations with Poland on a potential joint application, recognising that the compute density and investment scale required for a genuinely competitive gigafactory may exceed what either country can mobilise independently. The Czech-Polish coordination discussion reflects a maturation in Central European AI policy thinking, from nationally-focused initiatives that compete with each other for the same EU funding toward collaborative approaches that pool political capital and industrial capacity toward shared infrastructure assets that benefit the entire region.
Hungary’s entry into the AI infrastructure space through the ParTec AG and 3D Lézertechnika Zrt. modular hyperscale AI data centre announced in 2025, accompanied by a solar park and energy storage facility as part of a European flagship project in Central Europe, adds a commercially-oriented infrastructure layer to the publicly-funded AI factory and gigafactory pipeline. The Hungarian-language AI model jointly developed by the Ministry of Innovation and Technology and OTP Bank represents a parallel effort to build AI capability specifically calibrated for the Hungarian-language context that global frontier models do not optimally serve. Estonia’s introduction of AI education in secondary schools in partnership with OpenAI and Anthropic, Slovakia’s establishment of the region’s first high school with a specialised focus on artificial intelligence, and Google’s PhD scholarship programme in Bulgaria mentored by DeepMind researchers collectively illustrate that the talent pipeline investment now underway across Central Europe is more diverse, more serious, and more embedded in educational institutions than the region’s fifteen to twenty percent adoption lag relative to Western Europe might suggest.
The Window Is Open, But Not Indefinitely
McKinsey’s conclusion about Central Europe’s AI opportunity is framed in historical terms that reflect a genuine pattern in the region’s economic development over the past three decades. Central Europe has navigated structural transitions before, including market liberalisation after 1989, the integration of its countries into the European Union in 2004, and the shift to global supply chains in the 2000s and 2010s. In each case, organisations that moved decisively captured disproportionate value relative to those that waited for conditions to become more certain before committing. Those that waited paid a far greater price to catch up. The AI transition follows the same pattern but on a compressed timeline, because the technology is advancing at a pace that makes the cost of each additional year of delay larger than the equivalent cost in prior structural transitions.
The market pricing evidence that the McKinsey authors cite makes this concrete in terms that financial decision-makers should find difficult to dismiss. Palantir, which spent years building its operating model around AI-driven decision-making and data integration, saw its stock rise more than tenfold over the same period, with revenue growing seventy percent year over year. The difference between the two trajectories was not primarily one of technology quality or market positioning. It was whether AI was built into the operating model or bolted on too late. For Central European enterprises in manufacturing, financial services, or retail, the risk is the slower version of this market repricing: not a sudden collapse but a gradual erosion of competitiveness that becomes apparent only when the gap is too wide to close through the normal pace of internal transformation.
The conditions to act are genuinely better now than they were even two years ago. AI solutions have matured across the specific use cases most relevant to Central Europe’s dominant sectors, from production planning optimisation in manufacturing to end-to-end contact centre automation in financial services. Delivery models have been proven through enough large-scale deployments that the organisational learning is no longer confined to a handful of early movers. The regional public infrastructure investment in AI factories, gigafactories, and talent pipelines is beginning to create the shared foundation that individual enterprises cannot build alone. The CEE Digital Coalition’s recommendation for a regional AI supercluster, a joint AI fund, and coordinated grant programmes for entrepreneurs investing in AI reflects a growing consensus that the collaboration dimension of Central Europe’s AI opportunity is as important as the competition dimension. The seven hundred billion euro prize that McKinsey has quantified will not be claimed by the region simply because the potential exists. It will be claimed by the organisations and countries within the region that move from pilots to systematic integration first, that redesign their domains around AI rather than layering AI onto unchanged processes, and that treat the compression of the technology’s advancement timeline not as a reason to wait for clearer signal but as a reason to act before the window that is open now closes to the advantage of those who arrived earlier.
