Data Governance Revolution: DataVision EMEA 2025 Reveals Three Critical Shifts in Financial Services

The EDM Council's DataVision EMEA 2025 brought together the data management community at UBS London to explore "AI for Data Management – Culture, Ethics and Best Practices." What emerged from the discussions were three fundamental shifts that are reshaping how organisations approach data governance and analytics in the AI era.

→ View DataVision EMEA 2025 Event Details

The Observability Imperative: From Governance by Policy to Governance by Evidence

The most striking theme throughout the day was the industry's recognition that traditional governance approaches are insufficient for today's AI-driven data complexity. Multiple speakers emphasised that observability is becoming the cornerstone of effective data management.

Lafir Thassim's presentation on data observability resonated particularly strongly, highlighting the fundamental challenge: you cannot manage what you cannot see. This was reinforced by Matthew Widick from Compare The Market, whose approach of linking data and model observability through knowledge graphs exemplifies the emerging "Data for AI, AI for Data" paradigm.

The message was clear: organisations need to move beyond hoping their governance policies work to actually observing whether they deliver measurable outcomes. This shift from governance by policy to governance by observation represents a fundamental evolution in how we approach data management maturity.

Standards Evolution: DCAM v3 and ADAC Address Real-World AI Requirements

Figure 1: DCAM v3 Evolution - The EDM Council's updated Data Capability Assessment Model addresses AI-era requirements with enhanced Change & Enablement, Data Education, and Business Data Knowledge capabilities.

The presentation on DCAM v3 revealed significant evolution in how industry frameworks are adapting to AI-era requirements. The new capabilities around Change & Enablement, Data Education, and enhanced Business Data Knowledge reflect the framework's response to practical implementation challenges in AI-driven environments.

Figure 2: DCAM Measurement Framework - The shift from measuring 'maturity' to assessing 'readiness' represents a fundamental change in how organisations evaluate their data governance capabilities.

Particularly noteworthy was the shift from measuring "maturity" to assessing "readiness"—the framework now evaluates an organisation's fitness to perform specific capabilities rather than just their current state. This distinction matters enormously when implementing data governance in dynamic, AI-driven environments where agility and adaptability are paramount.

Figure 3: ADAC Framework Components - The AI, Data & Analytics Controls framework systematises AI governance with focus on automated key controls for ethics, privacy, and performance at enterprise scale.

The ADAC (AI, Data & Analytics Controls) framework developments demonstrate how the industry is systematising AI governance. The focus on automated key controls for ethics, privacy, and performance reflects a maturing understanding of what's required for scalable AI governance at enterprise level.

Figure 4: Automated Key Controls - Implementation of automated controls for AI, data, and analytics capabilities enables scalable governance whilst maintaining regulatory compliance and operational effectiveness.

We're witnessing the convergence of both domains (Data + AI) and capabilities (Governance, Management, and Observability)

Oli Bage's presentation using Solidatus demonstrated how traditional boundaries are dissolving. His visualization of LLM data lineage showed how data governance must now integrate with AI governance, how data management practices must evolve to support AI management needs, and how data observability capabilities must extend into AI observability. The ADAC framework exemplifies this multi-dimensional convergence, offering practical approaches that financial institutions can implement to navigate both innovation imperatives and regulatory requirements.

The Federated Reality: Balancing Scale with Quality

Catarina Palavichini dos Santos from UBS's Global Data Management Office provided valuable insights into the practical challenges of federated data management. Her evolution from requesting data investment to enabling AI platforms reflects a fundamental shift in how organisations view data value—from cost centre to strategic enabler.

The federated approach is clearly winning across financial services, but as the discussions highlighted, maintaining data quality at scale remains a significant challenge. This reinforces the critical importance of observability mechanisms that can provide consistent quality metrics across distributed data architectures.

The BCBS 239 Reality Check: A Decade of Persistent Challenges

Zak Watson's discussion of BCBS 239 implementation challenges provided a sobering reminder of the gap between regulatory requirements and actual implementation success. After a decade of industry efforts, the disconnect between policy-based governance and measurable compliance outcomes remains a critical challenge across financial services.

This regulatory reality underscores why observability frameworks are becoming essential. Financial institutions need evidence-based approaches to demonstrate compliance rather than relying on governance documentation alone.

Key Takeaways for Data Leaders

1. Observability is no longer optional. Organisations must invest in capabilities that provide real-time visibility into data quality, lineage, and governance effectiveness. The era of "governance by hope" is ending, particularly as AI adoption accelerates.

2. Standards are evolving rapidly. Frameworks like DCAM v3 and ADAC are adapting to AI requirements, but organisations need to stay current with these developments to remain competitive and compliant.

3. Convergence of both domains (Data + AI) and capabilities (Governance, Management, and Observability). Organisations that recognise this overlap gain a significant competitive advantage. Success in the AI era requires a unified approach that integrates governance, management, and observability capabilities across both data and AI domains.

Looking Forward: The Integration Imperative

The conversations at DataVision reinforced that integrated approaches connecting People, Process, Technology, and Data are essential for success in the AI era. The industry is clearly moving towards observability-driven governance, but implementation requires careful orchestration of organisational capabilities.

For financial services organisations particularly, the regulatory pressure combined with AI adoption requirements makes this shift from traditional to observational governance not just beneficial, but essential for competitive survival.

As the industry continues to grapple with these challenges, events like DataVision provide valuable forums for sharing practical experiences and evolving best practices. The question is no longer whether to implement observability capabilities, but how quickly organisations can make the transition whilst maintaining regulatory compliance and operational stability.

The EDM Council's DataVision EMEA series continues to provide essential industry dialogue on the evolution of data management practices. These conversations directly inform how we approach data and information observability frameworks and their practical application in regulated environments.

Thanks to the EDMC for hosting such a great event.