The focus on data quality will need to get stronger if pension administration is to capture the opportunities presented by artificial intelligence (AI), the Pensions Administration Standards Association (PASA) has argued.

In its newly launched industry guidance, Use of AI in Pensions Administration – Embrace the Opportunity with Caution, PASA described data as the “foundation of AI”, providing the essential information systems need to learn, make predictions and deliver useful insights.

The association stressed that data quality is “crucial” because it directly affects how well AI works and its accuracy, and that good, diverse data helps AI understand complex patterns and situations.

It added that high-quality data ensures that AI models are trained accurately, leading to better predictive capabilities and more reliable outcomes.

PASA Data Working Group chair, Kristy Cotton, warned that despite AI having “huge potential” to enhance efficiency and member experience, its “value depends entirely on the quality and integrity of the data behind it”.

“It’s about being curious, with caution, and ensuring AI works for the benefit of savers and the industry as a whole,” Cotton added.

In addition, PASA stressed that before trustees start implementing AI-led processes, they should make sure they have high confidence in their scheme’s members and governing data, with in-depth quality analysis and strong data management.

Although AI offers “transformational opportunities” for scheme administration, PASA also warned that the risks are significant and must be carefully managed.

In particular, it emphasised that continuous improvement, regular evaluation, and proactive risk management will be “crucial” in navigating the evolving landscape of AI in pensions administration.

While operational efficiency remains the primary initial focus of AI in pensions administration, PASA suggested that as AI technology - along with familiarity and trust in it - continues to evolve, its role within the pensions sector is expected to expand, offering more widespread solutions and enhancements.

PASA chair, David Fairs, noted that AI is already transforming the way administration operates and interacts with members, but it brings “real challenges” in data governance, ethics and transparency.

Given this, he explained: “This guidance helps schemes and administrators navigate this balance, embracing innovation while ensuring data quality, trust, and robust oversight remain paramount.”

The guidance intends to support the direction outlined by The Pensions Regulator’s Digital, Data and Technology Strategy and the Financial Conduct Authority’s AI Update, which both emphasise responsible innovation and robust data management.

---