An analytical commentary based on PwC Singapore’s Looking for ROI from AI? Focus on Scale
Can an Asian Report Explain Europe’s AI Dilemma?
PwC Singapore’s recent report, AI ROI: Focus on Scale (hereafter “the PwC report”), puts forward a clear argument: the key to realizing returns from AI investment is not whether companies adopt AI, but whether they can scale it. Organizations that successfully move from pilot projects to enterprise-wide deployment achieve revenue and efficiency gains up to 7.2 times higher than their peers.
This conclusion is compelling in the context of Singapore and the broader Asia-Pacific region. But when viewed from Europe—where the EU AI Act, GDPR, fragmented markets, and distinct innovation cultures define the landscape—the question becomes more complex. Does this logic still hold?
The answer is both yes—and not entirely.
I. A Shared Constraint: Europe Faces the Same Scaling Bottleneck
The PwC report highlights a widespread challenge: most companies experiment with AI, but few achieve meaningful scale. This pattern is just as evident in Europe.
A 2025 Deloitte survey of 1,854 executives across Europe and the Middle East confirms the same structural issue. While 85% of organizations increased AI investment in the past 12 months and 91% plan to continue, achieving satisfactory ROI typically takes two to four years—far longer than the 7–12 month payback period expected from other technologies. Only 6% of projects deliver returns within a year.
This aligns closely with PwC’s diagnosis: investment is rising, but value realization lags. The root cause is not the technology itself, but the lack of scaling capabilities—weak data foundations, insufficient workflow redesign, and underdeveloped governance structures.
Accenture’s research reveals further divergence within Europe. While 48% of large enterprises report having scaled strategic AI, only 31% of SMEs have done so. Eurostat data shows that, as of 2024, just 13.5% of EU companies use AI, with adoption ranging from nearly 28% in Denmark to only 3% in Romania.
Conclusion: PwC’s core insight—that scale is the key to ROI—applies to Europe as well, perhaps even more urgently.
II. The Structural Break: Regulation Changes the Game
This is where the PwC framework meets its limits. Singapore’s regulatory environment differs fundamentally from Europe’s.
Singapore promotes AI scaling through national strategies and regulatory sandboxes, with a clear emphasis on enablement and acceleration. Europe, by contrast, operates under the most comprehensive AI regulatory framework in the world.
The EU AI Act is being implemented in phases: prohibitions took effect in February 2025, obligations for general-purpose AI (GPAI) begin in August 2025, and full requirements for high-risk systems will apply from August 2026. Non-compliance can result in fines of up to 7% of global annual turnover—exceeding GDPR penalties. At the same time, companies must navigate overlapping regimes such as GDPR, DORA, and the Cyber Resilience Act.
A European Parliament study has already noted that while each regulation is justified individually, their combined effect may slow AI adoption, delay time-to-market, and create compliance asymmetries.
For European companies, this fundamentally alters the meaning of “scaling.” It introduces an additional layer: compliance as infrastructure.
Before redesigning workflows or replicating AI use cases, firms must classify risk levels, prepare technical documentation, and establish quality management systems. These are not strategic choices—they are legal obligations.
In Europe, PwC’s “three-step scaling model” requires a Step Zero: compliance readiness.
III. Data Governance: Europe’s Hidden Advantage
PwC highlights a clear gap between AI leaders and others in data capabilities. Among leading firms, 59% maintain a single source of truth, and 60% extensively use structured data, compared to 37% and 40% among Singapore respondents.
European companies face similar gaps—but with an important nuance.
GDPR has already forced many large European organizations to invest heavily in data governance: data classification, subject rights management, and processing records. While originally built for compliance, this infrastructure can serve as a foundation for AI.
In other words, Europe’s data governance baseline may be stronger than it appears—it simply has not yet been fully leveraged for AI.
IV. Cross-Industry Competition: Fragmentation Limits Replication
One of PwC’s most striking findings is that 43% of Singaporean firms are using AI to compete beyond their core industries, compared to a global average of 20%.
This dynamic is difficult to replicate in Europe—not due to lack of ambition, but because of structural constraints.
The EU consists of 27 member states with different languages, legal systems, and consumer behaviors. For a German company, expanding into France or Poland can be more complex than cross-industry expansion in Singapore.
Moreover, Europe’s economy is fundamentally SME-driven. Small and medium-sized enterprises account for roughly two-thirds of employment, yet they are the least equipped to scale AI. The World Economic Forum notes that European SMEs prioritize plug-and-play solutions with clear ROI, rather than large-scale model development.
V. Governance: From Compliance Burden to Competitive Asset
PwC identifies responsible AI frameworks and cross-functional governance as hallmarks of leading organizations.
Interestingly, Europe already leads in this area—but largely by necessity.
The EU AI Act mandates quality management systems, continuous risk assessment, and system registration for high-risk AI. According to a 2025 IAPP survey, 77% of European organizations are building AI governance frameworks, rising to nearly 90% among those already using AI.
This creates a unique opportunity. European firms can transform compliance into trust-based competitive advantage, particularly in regulated sectors such as finance, healthcare, and public procurement.
But doing so requires a shift in mindset: governance must be treated not as a cost center, but as a value driver.
VI. What Europe Should Take from PwC
Several key lessons from the PwC report remain highly relevant for Europe:
1. Stop treating the number of pilots as a KPI.
European firms often mirror the same mistake—running numerous AI experiments without mechanisms for scaling success.
2. Unlock the value of existing data governance.
GDPR-driven capabilities should be actively integrated into AI strategies, rather than operating in parallel silos.
3. Align compliance and scaling strategies.
Initiatives such as regulatory sandboxes, AI Factories, and the EU’s Apply AI Strategy provide structural support for scaling within a regulated environment.
4. Redesign workflows, not just add AI tools.
PwC notes that 56% of AI leaders fundamentally redesign workflows, rather than layering AI onto existing processes. This insight is equally critical for Europe.
The Framework Travels, the Execution Does Not
PwC’s central thesis—that scaling, not experimentation, drives AI ROI—holds true in Europe, perhaps with even greater urgency.
However, applying the framework requires three key adaptations: embedding compliance as a prerequisite to scaling, acknowledging the constraints of market fragmentation, and fully engaging SMEs as the backbone of Europe’s AI transition.
Europe is pursuing a fundamentally different AI trajectory. It is not optimizing for speed or risk tolerance, but for governance quality and trust.
PwC provides the map. Europe must build its own navigation system.
Sources:
PwC Singapore (2025), Looking for ROI from AI? Focus on Scale;
Deloitte (2025), AI ROI: The Paradox of Rising Investment and Elusive Returns;
World Economic Forum (2025), A New AI Playbook in Europe;
EU AI Act (official documents);
Lenovo/IDC, CIO Playbook 2026;
European Commission (2025), Apply AI Strategy.
Subscribe to EuroLuminant for independent European journalism.
