DataIQ

In a recent exclusive DataIQ peer-to-peer discussion, senior data and AI leaders in banking explored how organisations are redesigning governance, where regulatory expectations, model risk management, and third-party dependencies make governance more complex and more business-critical. The conversations highlighted practical ways to modernise existing approaches, using automation to improve efficiency while preserving human accountability.

Join an exclusive peer discussion for senior data and AI leaders. This small-group session offers a confidential space for senior professionals to exchange insights, share challenges, and learn from one another. ────────────────────────────────────────────────────── Discussion Questions: How are organisations extending governance beyond traditional critical-data domains to support AI, and what approaches are proving most effective?

By Gaurav Rastogi, Senior Director – Data Analytics & BI Platforms, Hertz Disclaimer This article reflects the personal views and professional experiences of the author. It is based on original frameworks, methodologies, and observations developed through enterprise-scale data platform transformations. The perspectives shared do not represent any specific employer or affiliated organization. Any references to technologies or platforms are for illustrative purposes only and do not imply endorsement.

Assess your current architecture and make informed choices about what to modernise, replace, or integrate as demand for AI-ready infrastructure accelerates. This event is designed for DataIQ clients only. If you’re a client, you can sign up via the DataIQ Hub. If you are not a client yet, you can find our more here. ────────────────── Discussion questions: Where are the biggest points of friction in your current data and analytics architecture?

The biggest change is a delay to obligations for high-risk AI systems. Standalone high-risk systems covered by Annex III, including those used in areas such as employment, education, credit scoring, and law enforcement, will now be subject to the rules from 2 December 2027 instead of 2 August 2026. High-risk AI embedded in regulated products, such as medical devices, will have until 2 August 2028 to comply. Watermarking requirements for some AI systems are also postponed until December 2026.

Most large companies, from financial services to retail, have made AI personal. Tens of thousands of assistant licences have helped employees save hours a day delivering real and measurable uplifts in individual productivity. Yet almost none of this has translated into strategic value: Cost-to-income has not moved. The combined ratio has not moved. The numbers the board actually report on are unchanged.

The discussion moved quickly beyond identifying use cases. Instead, it focused on the practical realities of managing competing demands—balancing foundational investment with pressure for short-term results, filtering an influx of AI-driven ideas, and navigating the persistent difficulty of proving value. What emerged was a set of working principles for structuring and governing a portfolio that can both deliver now and scale over time.

The full insights are available exclusively to DataIQ subscribers. Data and AI leaders with a focus on governance and risk from financial services, insurance, retail, manufacturing, consumer goods, government, and energy organisations shared how they are responding to the growing complexity of AI adoption.

The emergence of data products has changed the equation. As organisations seek to create reusable, governed, and discoverable data assets, data models are increasingly serving as the semantic foundation that connects business concepts, physical data assets, governance processes, and AI-driven consumption. Against this backdrop, a growing number of organisations are exploring how GenAI can accelerate data modelling, improve consistency, and unlock new ways of interacting with enterprise data.

This article is written in partnership with Rishi Kapoor, Head of WW Partner Sales Engineering, Alteryx & DataIQ 100 2026 Enabler. AI experimentation has become easier, but AI production remains an operating commitment. By some estimates cited by RAND, more than 80% of AI projects fail.