AI has value when it is woven into everyday work

Technology that makes analyses faster and smarter seemed very promising to him. At the same time, he had doubts about its reliability, especially in a sector where accuracy and trustworthiness are so important. He saw great potential, but only if it was applied responsibly and with careful consideration.

AI as part of digitalisation
Within PGGM’s Policy Advisory department, AI is not a standalone effort but part of a broader digitalisation programme. “AI is not a separate initiative for us,” Erik explains. “It is part of the way we want to work smarter, more efficiently, and in a future‑proof way.”

His team is renewing advisory products with AI‑supported analyses, helping colleagues become more digitally skilled through learning communities and workshops, using modern data platforms such as DDI (Data Driven Infrastructure) and DAP (Data Analytics Platform), and ensuring the secure integration of AI into the IT architecture. Processes are also being further automated with AI, RPA (Robotic Process Automation), and smart workflows. Erik says: “From that foundation, we work on practical applications, from document analysis to experiments with generative AI. Not because we have to, but because it genuinely delivers results.”

AI integrated into everyday work
According to Erik, AI only has real value when it is truly woven into the work. That is why, with the Future Advise programme, they focus not only on technology but also on skills, structure, and engagement. “The outlier analysis helps us automatically detect irregularities in the pension administration at StiPP. We will soon do the same for PFZW and Schilders,” he says. “This improves data quality and enables us to deliver more reliable products, such as annual reports or policy advice.”

Other teams are also working with concrete AI applications. “The Case Law Robot supports colleagues in Pension Legal Affairs and Policy with file analysis and complaints handling. And the sickness absence model for employers in the health and welfare sector, which we developed a few years ago, remains extremely valuable. It predicts long‑term absenteeism and identifies risk factors. It helps employers take preventive action, which fits perfectly with our role in themes such as Living Vital and Prepared for Retirement.”

In addition, more and more employees are being supported by smart AI tools. GitHub Copilot helps modellers write and improve code. Policy advisers work with an AI‑driven knowledge base. Smarter forms of complaints handling are being implemented. “We also use AI to monitor data quality and to analyse participant behaviour and pension outcomes,” Erik adds. “AI is not an extra, it is becoming a fixed part of how we work.”

Learning, experimenting, and collaborating
The most important lesson has been that AI is not just about technology. “It’s about colleagues understanding how AI can make their work easier, smarter, or more enjoyable,” says Erik. “That’s why we provide tailored training within teams, explaining tools such as the PGGM Intelligent Agent and Microsoft 365 Copilot in simple, practical terms: what does it do, and how can it help you?”

The ethical side is also given attention. “We organised a workshop on the legal and ethical aspects of AI. Technology is wonderful, but we want to use it carefully and transparently. What truly makes it successful is curiosity, the space to experiment, and collaboration between subject‑matter experts and data specialists. AI therefore requires not only technical knowledge but also new skills, such as effective prompt use, critically assessing AI output, and knowing when to use a tool, and when not to. That confidence grows a little more every day.”