ING’s AI Roadmap: Platform, People, and Agentic AI

Marco Li Mandri sets out ING’s three-part AI plan for the next two to three years: move fast, scale responsibly, and make sure people can actually use the tools. The goal isn’t scattered pilots, but AI becoming part of everyday work in a few big-impact areas, backed by the right controls and skills.

First, Li Mandri says ING wants to rapidly scale AI and agentic AI in focus domains like the contact centre and wider customer service. Practically, that means changing how customers get help and how staff deliver it, not just answering FAQs, but summarising issues, suggesting next steps, pulling relevant information quickly, and routing cases while also flagging retail customer interaction, where AI could smooth high-volume journeys like onboarding, product questions, payments, disputes, and fraud, making experiences more personalised and proactive.

ING also describes building one global platform to serve all countries and domains, concentrating ING’s agents and technology in a shared foundation which supports faster reuse across teams and makes the essentials easier to standardise: model onboarding, monitoring, auditability, data access, and change management. It also strengthens safety and responsibility through consistent guardrails, clearer accountability, and tighter control over data access and output monitoring which is critical in a regulated, high-stakes environment.

Lastly, Li Mandri focuses on people as leaders need to learn how to operate in an AI-driven world, including realistic capabilities, risks, and how work changes when humans and AI share tasks. Colleagues need practical skills too: prompting well, judging output quality, and staying critical. ING’s data fluency programme is meant to build these habits across the organisation, reinforcing that AI strategy is as much about people and governance as it is about models and platforms.