From cloud adoption to business value
Cloud transformation has improved scalability, resilience and agility through infrastructure and application modernisation. Yet data often remains fragmented across public, on-premises and sovereign cloud environments, limiting its use in business processes.
The next step is to build on these capabilities to make data more accessible and reliable, enabling better decisions and more efficient operations.
Data platforms become a strategic asset
According to Gartner, data management software spending will reach $21.6 billion by 2029, growing at a 9.3% CAGR, driven by demand for integration and governance. These foundations are also critical for AI: Gartner predicts that, through 2026, 60% of AI projects lacking AI-ready data will be abandoned.
Most organisations already have plenty of data. Customer and business information is spread across CRM systems, reference data stores, data warehouses, billing platforms, support and operational systems, and End-User Computing tools (EUCs). This fragmentation creates inconsistencies, operational overhead and risk. Cloud migration moved these silos; it rarely merged them.
A modern data platform brings information together, applies consistent quality rules and access rights, and makes it reusable for reporting, operations and AI. Success depends on clear ownership and quality standards.
The platform must also stay lean: every new tool adds integrations, data flows and costs.
As AI models become more accessible, an organisation’s own data and business knowledge become key differentiators. A strong data foundation helps organisations capitalise on this advantage.
AI raises the stakes
Testing a model takes an afternoon. Running it reliably inside a business process is another matter, especially with AI agents that read documents, query systems and act on what they find.
Agents depend on data. Without clear ownership, known quality and defined permitted uses, they cannot be trusted or deployed. Poor governance can also expose information to unauthorised people.
Scaling AI also brings cost challenges: while token prices have generally fallen, exponential growth in usage can increase spending. Organisations need to monitor consumption and manage costs as AI scales.
Sovereignty: control without isolation
As reliance on data and AI grows, sovereignty is key to maintaining control over data, operations and technology.
Keeping data in Europe is not enough. Organisations must also consider operational control and technology dependencies.
According to PwC’s 2025 EMEA Cloud Business Survey, 84% of organisations in Luxembourg are refining their cloud strategies to address sovereignty.
Data platforms enable analytics and AI but can deepen dependencies. Portability, interoperability and exit options must be considered from the outset.
For financial institutions, the Digital Operational Resilience Act (DORA) requires credible exit strategies for services supporting critical functions. Exits may involve applications, reporting and business processes, not just data. Open formats help, but critical workloads must be tested for portability, including processing logic and dependencies.
Sovereignty does not mean avoiding major cloud providers or moving everything to sovereign infrastructure. Each cloud model involves trade-offs. The aim is to match control to workload sensitivity and criticality while retaining flexibility to innovate and change providers.
Making cloud and data work together
Matching control to each workload is easier said than done. Data often remains organised by business function, with finance, sales and operations managing their own information. Each new cloud service can add to this fragmentation.
Progress starts with a simple question: who is responsible for this data? When a business owner is accountable for its quality and use, data can be shared with confidence. Common integration and security rules allow it to move between applications and cloud environments instead of being copied.
This takes time. Programmes that address all data issues at once tend to stall. A better approach is to integrate data one business area at a time, starting with valuable use cases, increasingly involving AI.
The next stage of cloud transformation
As AI moves from pilots to production, organisations must adapt cloud environments to support AI at scale. Existing cloud architectures must evolve to integrate AI services such as Azure OpenAI and Amazon Bedrock into business applications, with the right governance, security and cost controls over data access, model usage and AI agents’ actions. They must also ensure required AI services are available in the right regions and can meet demand.
Finally, organisations must track AI usage and costs as adoption grows. The goal is sustainable business value without losing control over security, performance and costs.
Cloud, data and AI transformation is creating new opportunities for organisations. Explore PwC’s perspective on the evolving cloud landscape.