TRENDS AND INSIGHTS The AI-Ready Data Gap: Why Data Enrichment Powers Agentic Banking

AI ambition is accelerating, but the data foundation is lagging

Artificial intelligence investment across financial services is accelerating rapidly, with spend projected to rise from $31 billion in 2024 to $81 billion by 2028¹. Organisations are moving beyond experimentation with AI agents into deployment: 50% of organisations report they are already deploying AI agents in production across multiple business areas, and a further 27% have AI agents in production within one business area².

The ambition is clear: AI is no longer a future capability, it is becoming the engine of modern financial experiences.

Yet a critical gap remains: a significant amount of transaction data is not AI-ready.

Transaction data sits at the heart of financial services, but it remains one of the most complex and least standardised datasets. In practice, this can result in data that is inconsistent, fragmented across systems, and difficult to interpret at scale. This creates a fundamental mismatch: issuers are trying to power next-generation AI on weak data foundations. 

Why this matters for issuers now: AI changes the stakes

In the era of agentic AI, the quality of an AI system is only as strong as the data behind it. Unlike traditional analytics, AI systems do not just analyse data – they act on it.  

AI systems depend on clean, structured, and reliable data. Without this foundation- and where they are forced to operate on inconsistent information - the risk of poor customer experiences, incorrect recommendations, compliance issues, and operational inefficiencies is increased. As a result, data quality becomes a control point for risk, trust, and performance.

At the same time, customer expectations are increasing rapidly- 62% of consumers say they trust banks to deliver generative AI services.³

Closing the gap: from raw data to AI ready intelligence 

To unlock the full value of AI, issuers must address this issue at its source: foundational data quality.

This is where solutions such as Visa’s Data Enrichment can help address this challenge. At its core, Data Enrichment transforms raw transaction data into cleansed and structured data, providing the foundation for agentic products. The solution transforms messy, fragmented data into a 360-degree understanding of spend:

Transforming raw transaction data into structured, enriched insights creates value for both consumers and issuers: customers can benefit from clearer and more intuitive experiences whilst issuers can gain a more reliable foundation for AI innovation and agent-driven decision making.

In practice, Data Enrichment can help issuers to: 

• Transform messy transaction data into clear insights with recognisable merchant details;

• Make data usable and visually clear from day one, reducing the need for manual cleanup; and 

• Help create a more unified view of transaction data to improve AI accuracy and decision speed.

Building the foundation for what comes next

As AI adoption accelerates among issuers, the value generated by AI will increasingly depend on the quality of the data that powers it. The key question for issuers is no longer “Do we have an AI strategy?”, it is “Do we have the data foundation required to support AI at scale?” 

Transforming fragmented transaction data into a trusted, structured foundation can enable issuers to unlock more accurate insights, better customer experiences, and greater value from AI. Data enrichment helps build the foundation for AI‑ready transaction data and agent-enabled banking experiences. 

To learn more about how Visa can help you build an AI-ready data foundation through Visa Data Enrichment, reach out to your Visa representative.

This content was developed for a European audience and relates to Visa’s offerings in Europe