TWINLADDER
TwinLadder
TWINLADDER
Subscribe
Back to Case Studies

Twin Ladder Casebook

The Ambition Gap

February 28, 2026|media report

Somewhere in a glass-walled conference room in Rotterdam or Munich or Lyon, a board meeting is underway. The strategy deck is polished. Slide fourteen bears the heading "AI-First by 2027," and the room nods approvingly.

The Ambition Gap

Listen to this article

0:000:00

The Ambition Gap

Why Europe Leads in AI Plans and Lags in AI People

Twin Ladder Casebook --- Twin Ladder, February 2026

The Hook

Somewhere in a glass-walled conference room in Rotterdam or Munich or Lyon, a board meeting is underway. The strategy deck is polished. Slide fourteen bears the heading "AI-First by 2027," and the room nods approvingly. The Chief Digital Officer walks the board through the investment: new platforms, new partnerships, a new Centre of Excellence. The numbers look right. The ambition is real.

Then someone asks a quieter question. How many of our people have actually been trained to work with these systems?

The room goes silent. The answer, across European enterprises, is devastating. A 2024 survey of 1,400 executives found that more than eighty-six percent of European companies rank AI investment as a top-three priority --- placing Europe ahead of every other region in stated ambition. Yet fewer than one in twenty of those same companies have trained more than a quarter of their workforce to use AI in any meaningful capacity. The strategy deck says AI-first. The shop floor says AI-absent. The distance between the slide and the reality is not a gap. It is a chasm, and it runs across the entire continent.

The Story

The European ambition gap is not a single failure. It is a mosaic of national failures, each shaped by local conditions and all producing the same result: organizations that have committed to AI in strategy while neglecting the humans who must operate it.

The Netherlands: Infrastructure Without People

The Netherlands is, by most measures, one of Europe's best-positioned economies for digital transformation. It ranks among the top nations in broadband penetration, cloud adoption, and venture capital per capita. It hosts Booking.com, ASML, and a dense cluster of AI startups around Eindhoven and Amsterdam. On paper, Dutch enterprises should be leading Europe in AI deployment.

The Cisco AI Readiness Index tells a different story. In its 2024 assessment of 7,985 senior business leaders across thirty markets, Cisco found that only thirteen percent of organizations globally are fully prepared to deploy AI --- down from fourteen percent a year earlier. The Netherlands, despite its digital infrastructure, scored near the bottom of the readiness spectrum. Dutch respondents reported acute shortfalls in the talent pillar: only thirty-one percent of organizations globally claimed high talent readiness, and Dutch enterprises fell below even that modest benchmark. The country has built excellent pipes. It has not filled them with capable people.

The pattern is instructive. The Netherlands invested heavily in the hardware of AI --- connectivity, compute, data centres --- while treating the software of AI, the human competence to evaluate, direct, and challenge AI output, as an afterthought. The result is a nation where the infrastructure awaits a workforce that does not yet exist.

France: Ambition Meets Anxiety

France presents a different facet of the gap. The Macron government has positioned AI as a national strategic priority, committing billions to research and announcing plans to make France the leading AI hub in continental Europe. The rhetoric has been forceful, the investment significant.

The reality at the enterprise level is sobering. According to Eurostat data, only about ten percent of French companies were using AI technologies in 2024 --- a mere four-point increase from the year before. Meanwhile, an IPSOS survey published in early 2025 found that seventy-three percent of French citizens fear AI's impact on society, the highest level of concern across a global panel of eleven countries. France has the rare distinction of combining high governmental ambition with low enterprise adoption and elevated public anxiety. The population does not trust AI. The companies have not adopted it. And the government is pushing forward regardless.

The fear is not irrational. French labour markets are shaped by strong worker protections, a deep tradition of social dialogue, and legitimate concerns about technological displacement in a country where youth unemployment already hovers near twenty percent in certain demographics. But fear without competence is paralysis. French workers are not being given the skills to evaluate whether their anxieties are well-founded, to distinguish between AI applications that threaten their roles and those that enhance them. The absence of AI literacy does not prevent AI from arriving. It prevents people from shaping how it arrives.

Germany: The Training Desert

Germany, Europe's largest economy, demonstrates the gap at industrial scale. By mid-2025, roughly forty-one percent of German companies reported using AI in some capacity --- a significant figure, and well above the EU average of twenty percent. Germany's manufacturing base, its Mittelstand tradition, and its engineering culture have produced genuine enterprise adoption.

But adoption without training produces dependence, not competence. Germany's AI deployment has outpaced its AI education. The country's dual vocational training system, long admired as a model for workforce development, has been slow to integrate AI literacy into its curricula. The result is a workforce that increasingly uses AI tools without understanding how to evaluate what those tools produce. This is not a theoretical concern. It is the precise condition that the Competence Paradox describes: tools that accelerate individual output while quietly degrading the judgment required to direct that output wisely.

The Nordics: Slipping From the Top

Perhaps the most alarming data comes from the nations that should be leading. The Nordic countries --- Denmark, Finland, Sweden, Norway --- have long occupied the top tier of global digital readiness rankings. They have invested heavily in digital government, broadband infrastructure, and education technology. They are small, wealthy, well-governed, and technologically literate. If any region should be able to close the ambition gap, it is the Nordics.

The rankings tell a different story. Finland's position in the Global AI Index declined from tenth in 2023 to fifteenth in 2024. Denmark slipped from sixteenth to twenty-second. These are not catastrophic falls, but they represent a trend: nations that once led in digital readiness are losing ground in AI readiness, precisely because AI demands a different kind of capability than the digital transformation that preceded it.

The EY Work Reimagined Survey of 2025 quantified part of the problem: in Sweden, sixty percent of employees reported receiving fewer than four hours of AI training. Only thirteen percent had received more than forty hours. The Nordics have digital infrastructure. They do not yet have AI-competent workforces. The gap between the two is the ambition gap in miniature.

The Continental Picture

Eurostat's 2025 enterprise survey places the EU-wide AI adoption rate at twenty percent of enterprises with ten or more employees --- up from 13.5 percent in 2024, but still a figure that means four out of five European enterprises have not adopted AI at all. Among those that considered and rejected AI adoption, the most commonly cited barrier was a lack of relevant expertise, reported by seventy-one percent of respondents. Not cost. Not regulation. Not infrastructure. People. The binding constraint is human.

Through the Twin Ladder Lens

The Twin Ladder framework, introduced in The Competence Paradox (Twin Ladder, 2026), describes a four-level progression for building AI competence: Level 0 (AI Literacy Foundation), Level 1 (Professional Twin), Level 2 (Operational Twin), and Level 3 (Ecosystem Twin). Each level builds on the one below. The ladder is climbed, not skipped.

Europe's ambition gap is, in Twin Ladder terms, a Level 0 crisis. The continent is attempting to operate at Levels 2 and 3 --- deploying ecosystem-scale AI platforms, building digital twins of supply chains, modelling entire operational environments --- without having established the foundational literacy that makes those deployments meaningful. The retail chain described in Chapter 7 of the whitepaper, which built a technically impressive ecosystem model that nobody in the organization could evaluate, is not an outlier. It is the European norm.

Level 0, AI Literacy Foundation, is defined as the baseline ability to critically assess AI-generated output. It means knowing that a confident, well-structured response from a large language model can be factually wrong. It means understanding that an AI recommendation is a statistical inference, not a verified conclusion. It means recognizing the difference between output that is useful, output that is plausible but flawed, and output that is dangerous.

This is precisely the capability that European workforces lack. When seventy-one percent of European enterprises cite "lack of relevant expertise" as their primary barrier to AI adoption, they are describing a Level 0 deficit. When Dutch enterprises score poorly on talent readiness despite world-class infrastructure, they are describing a Level 0 deficit. When French citizens fear AI because they cannot evaluate what it does, they are describing a Level 0 deficit.

The EU AI Act recognizes this. Article 4, which entered into force on 2 February 2025, requires that providers and deployers of AI systems "take measures to ensure, to their best extent, a sufficient level of AI literacy of their staff and other persons dealing with the operation and use of AI systems on their behalf." The obligation is not limited to technical operators. It covers anyone involved in the operation and use of these systems. Non-compliance is treated as an aggravating factor when assessing penalties for other AI Act violations. From August 2025 onward, providers and deployers who fail to ensure adequate literacy face potential civil liability if inadequately trained staff cause harm.

Article 4 is, in effect, a legal mandate for Level 0. The European Union has legislated the bottom rung of the Twin Ladder. And yet most European organizations are nowhere near compliance. They have strategy decks about AI. They do not have workforces that can evaluate what AI produces.

You cannot build Professional Twins on a workforce that has not achieved basic AI literacy. You cannot run Operational Twins when your people cannot distinguish between a statistically robust recommendation and a plausible hallucination. You cannot model ecosystems when the decision-makers reading the models lack the foundational competence to challenge what they see. The ladder is climbed, not skipped. Europe has not started climbing.

The Pattern

The European ambition gap sits atop a deeper, more troubling foundation. The OECD's Survey of Adult Skills (PIAAC), published in December 2024, found that literacy and numeracy skills among adults in most OECD countries have stagnated or declined between 2012 and 2023. This is not a marginal finding. It is a decade-long erosion of foundational cognitive capabilities across the developed world, and it strikes directly at the preconditions for AI literacy.

The decline has not been evenly distributed. The OECD data reveals that low-educated adults experienced larger and more widespread declines in both literacy and numeracy than their better-educated counterparts. The literacy proficiency of the lowest-performing ten percent of the population has fallen in most participating countries. The result is a widening skills inequality: a gap between educated and uneducated adults that was already substantial in 2012 and has grown larger over the following decade.

This matters for AI literacy because the people most likely to be excluded from AI training are the people who need it most. Current training systems, as the OECD has documented, favour those already advantaged by higher education. The information barriers, time constraints, and cost structures of professional development programmes systematically exclude lower-skilled workers. When European enterprises do invest in AI training, the investment flows disproportionately toward knowledge workers --- the software engineers, data scientists, and product managers who are already closest to understanding AI. The warehouse supervisor, the shop floor technician, the administrative assistant --- the workers whose roles are most likely to be reshaped by AI --- are the last to receive training and the first to be displaced.

The ambition-competence gap is therefore not merely an enterprise problem. It is a societal one. Europe is building an AI-powered economy on a workforce whose foundational skills are declining, whose access to training is inequitable, and whose capacity to evaluate AI output has never been systematically developed. The strategy decks promise transformation. The skills data promises fracture.

This is the pattern the Twin Ladder is designed to interrupt. Level 0 is not an optional first step for early adopters. It is the necessary foundation for an entire workforce --- including, and especially, the workers who have been excluded from every previous wave of digital upskilling.

The Lesson

AI Literacy is not a nice-to-have. As of February 2025, it is a legal obligation under the EU AI Act. Article 4 does not distinguish between a technology company and a bakery chain, between a headquarters in Berlin and one in Bratislava. If an organization deploys AI systems within the European Union, it must ensure that the people operating those systems possess a sufficient level of AI literacy. The compliance obligation is live. The enforcement framework is active. And the vast majority of European enterprises have not begun.

The Twin Ladder begins at Level 0 for a reason. An organization that cannot evaluate AI output cannot govern AI deployment. It cannot identify when an AI system is producing biased recommendations, hallucinated content, or statistically unsound predictions. It cannot comply with Article 4. It cannot build the Professional Twins of Level 1, the Operational Twins of Level 2, or the Ecosystem Twins of Level 3. Without Level 0, the entire structure is built on a foundation of uncritical acceptance --- and uncritical acceptance is precisely what the Competence Paradox warns will erode human capability over time.

The path forward is not complex. It requires asking one question before any other: Can your people evaluate AI output? Not use AI. Not prompt AI. Evaluate it. Can they distinguish between a recommendation that is robust and one that is plausible but wrong? Can they identify when an AI system is operating outside its competence? Can they explain, in language their colleagues understand, why they accepted or rejected what the AI produced?

If the answer is no, then every AI investment the organization has made is unmoored. The strategy deck is aspirational fiction. The first investment is not in platforms or models. It is in people.

Monday Morning Question: If you removed every AI tool from your organisation for one week, which decisions would your people no longer know how to make --- and what does that tell you about where to start?

Sources

  1. Fortune Europe (2024). "Europe is eager to embrace AI --- but ranks worst in readiness." Survey of 1,400 executives on AI investment priorities and workforce training. https://fortune.com/europe/2024/01/30/europe-ai-readiness-training-workers/

  2. Cisco AI Readiness Index (2024). Global survey of 7,985 senior business leaders across 30 markets on AI deployment readiness, infrastructure, talent, data, and governance. https://www.cisco.com/c/m/en_us/solutions/ai/readiness-index.html

  3. Eurostat (2025). "20% of EU enterprises use AI technologies." Enterprise survey on AI adoption rates across EU member states, including barriers to adoption. https://ec.europa.eu/eurostat/web/products-eurostat-news/w/ddn-20251211-2

  4. EU AI Act, Article 4: AI Literacy (2024). Full text of the mandatory AI literacy provision, effective 2 February 2025, requiring providers and deployers to ensure sufficient AI literacy among staff. https://artificialintelligenceact.eu/article/4/

  5. OECD (2024). "Adult skills in literacy and numeracy declining or stagnating in most OECD countries." Results of the PIAAC 2023 Survey of Adult Skills, documenting decade-long declines in foundational cognitive skills. https://www.oecd.org/en/about/news/press-releases/2024/12/adult-skills-in-literacy-and-numeracy-declining-or-stagnating-in-most-oecd-countries.html

  6. IPSOS / French Tech Journal (2025). "French AI Paradox: Speed Up or Slow Down?" Survey data on French AI adoption rates and citizen attitudes, including the finding that 73% of French citizens fear AI's societal impact. https://www.frenchtechjournal.com/french-ai-paradox-speed-up-or-slow-down/

  7. EY Work Reimagined Survey (2025). "Sweden and the Nordic countries are lagging behind in AI use." Survey of 15,000 employees and 1,500 employers across 29 countries on AI training hours and workforce readiness. https://itbranschen.com/en/sweden-the-nordic-countries-are-lagging-behind-in-ai-usage/

  8. Twin Ladder (2026). The Competence Paradox: Why AI-Powered Organizations Need AI-Competent People --- and How to Build Them. Whitepaper introducing the Twin Ladder framework. https://twinladder.lv/whitepaper

This article is part of the Twin Ladder Casebook, a series by Twin Ladder examining how organizations across industries and regions confront the gap between AI ambition and AI competence. Each case is viewed through the Twin Ladder framework introduced in The Competence Paradox (2026).