Report: Data gaps are hindering AI adoption in banking
A new report from Info-Tech Research Group finds that while banks are accelerating AI initiatives, many are hitting a wall due to legacy data environments. Structural limitations and a lack of real-time data are reportedly the biggest hurdles to scaling AI in fraud detection, personalization, and risk analytics.
The core issue for banks is that legacy systems were built for stability and static reporting, not the dynamic, real-time data streams required by modern AI. Architectures designed around batch processing create risks and limit the ability to use predictive insights across different business units. A recent survey found that 68% of financial services CTOs see these legacy systems as the biggest barrier to AI adoption. This data problem is not just about volume, but also variety and velocity. AI models for fraud detection and personalization need to process unstructured digital signals and behavioral data at scale, which is a fundamental mismatch with the structured, historical data that legacy systems are designed to handle. This often leads to data being stored in fragmented silos, creating inconsistencies that can lead to inaccurate AI predictions. The push for AI adoption is significant, with global financial institutions expected to spend over 20% of all AI-related expenditures between 2024 and 2028. Banks that successfully use advanced AI models report fraud detection accuracy rates exceeding 90%. However, Deloitte's 2024 Financial AI Adoption Report found that only 38% of AI projects in the finance sector meet or surpass their expected return on investment, with over 60% of firms reporting significant delays in implementation. To overcome these hurdles, a strategic shift is required, moving from project-level integration to an enterprise-wide data strategy with strong governance. This involves modernizing data architecture to support AI-driven capabilities without compromising regulatory discipline. For many, this means creating unified data lakes, improving data quality, and leveraging cloud solutions to process vast amounts of information at scale. This trend is giving rise to "agentic workflows," where autonomous AI agents manage complex tasks with minimal human input by perceiving their environment, making decisions, and adapting in real time. Gartner predicts that by 2028, a third of all enterprise software applications will incorporate this type of agentic AI. For financial services, this could mean AI agents handling everything from customer inquiries to initial fraud analysis. In the UK, the tech sector remains a major European hub, with startups raising over $7 billion so far in 2025. London-based AI and fintech companies are attracting significant funding. For instance, AI company ElevenLabs raised £145 million in January 2025. This robust ecosystem provides a competitive landscape for talent and innovation as established banks race to modernize. The transition to CTO at a growth-stage SaaS company requires evolving from a hands-on technical expert to a strategic leader. The focus shifts from writing code to setting architectural direction, scaling engineering teams, managing budgets, and aligning technology with broader business goals. A key challenge is balancing the need to pay down technical debt while maintaining innovation velocity. On the Formula 1 front, the 2026 season is dominated by major technical regulation changes, particularly around engine power units. The new rules mandate a 50/50 power split between the internal combustion engine and electric motors, alongside a move to 100% sustainable fuels. These shifts are attracting manufacturers like Audi, Ford, and Honda back to the sport, seeing it as a key R&D platform for future road car hybrid technologies.
