AI as a Foundation for Future-Proof Pension Management

Artificial intelligence (AI) is evolving rapidly. For years, AI was largely the domain of specialists, but the rise of generative AI has suddenly made it accessible to a much broader audience. At PGGM too, colleagues started experimenting with AI applications, while business-driven use cases began to emerge across the organisation.

The challenge quickly became how to organise all of this in a responsible way. 

By the end of 2024, it was clear to me that Pension Management needed to take a more structural approach. Not through isolated experiments or one-off initiatives, but through a shared strategy built on solid foundations. Since then, we have been working step by step to embed AI as a lasting part of our services and organisation. 

Lasting change 

Where earlier technological revolutions mainly automated repetitive tasks, AI is now transforming knowledge work itself. That is equally true for us. Within PGGM, we saw colleagues increasingly incorporating AI into their daily work. At the same time, new legislation and regulations emerged, including the European AI Act.

This calls not only for innovation, but also for clear frameworks, governance and AI literacy across the organisation. That is why, at the start of 2025, we began developing an AI strategy for Pension Management, aligned with our strategic objectives. 

A solid foundation 

What we deliberately wanted to avoid was creating isolated AI solutions without any overall cohesion. LinkedIn is full of new AI demos every day, but we felt it was more important to get the foundations right first. That means investing in AI governance, data quality, architecture, infrastructure, AI literacy, and clear development and operational processes. Because without strong foundations, AI cannot be scaled responsibly. 

The implementation of our AI strategy focuses both on building those foundations and on delivering strategic use cases. 

Future perspectives 

Within our strategy, we work with four future perspectives: customer interaction, processes, employees and IT. These help bring to life how AI could reshape Pension Management in the years ahead.  

In customer interaction, for example, we expect participants and employers to increasingly communicate with us through AI-driven channels. Not as a replacement for existing channels, but as an addition to them. AI can help make complex pension information faster to access, more personalised and easier to understand.

We also see major opportunities within processes. We believe processes will become increasingly data driven and AI powered. By integrating AI agents intelligently into workflows, certain steps can be accelerated, giving employees more time to focus on work where human judgement and attention truly add value.  

This will also change the nature of work for employees. I do not expect teams to disappear, but I do expect employees to work more closely alongside AI agents. Their role will gradually shift from carrying out every task themselves to directing, reviewing and managing processes. Internally, we sometimes refer to this as the role of the ‘AI conductor’. 

We also see significant potential within IT, for example in documentation, test scenarios, code reviews and software development. In these areas, AI can take over a range of supporting tasks. 

Use cases 

Together with our appointed AI ambassadors from around the Pension Management organisation, we have now identified more than 150 use cases. These range from customer interaction and process optimisation to employee support. One example is the use of AI in onboarding new colleagues. In some teams, it takes considerable time to familiarise employees with all the relevant work instructions, policy documents and processes.

By making this information accessible through an AI agent, new colleagues can independently ask questions and instantly receive answers based on the correct source information. This reduces pressure on colleagues while speeding up the learning process. 

We have also established an AI Value Office, where colleagues can submit ideas and use cases. Together, we assess whether AI is genuinely the right solution, what is needed to make it work, and how a use case can be developed in a controlled way towards production. 

In addition, we are actively exploring the use of AI agents within our pension Management processes. To support this, we are investigating various agentic process automation platforms. The aim is to deploy AI agents within production processes in a controlled and manageable way, with appropriate monitoring, governance and human oversight where needed. This marks the next step in our AI maturity. 

Core competence 

Within our department, we are gradually helping to make AI a core competence for PGGM. Not because it is a hype, but because it is a technology that will fundamentally change our work, our processes and our services. 

In this article, we have focused mainly on the strategy, the foundations and the first applications within processes and employee support. In future blogs, we will explore concrete examples from the other future perspectives in more detail, including customer interaction, AI agents within our organization and the collaboration between employees and AI.