AI & Digital

AI is used to process large volumes of data in research, to monitor risks, to automate reporting and operations and to support client service. For asset owners, it also matters as an investment theme that runs across listed equities, private equity, infrastructure such as data centres, and energy.

Boards remain accountable for decisions supported by AI. Key questions concern model risk and explainability, data quality and protection under the revised Swiss Federal Act on Data Protection, outsourcing to technology providers, and clear responsibilities for monitoring how models are used.

FINMA has set out its supervisory expectations for supervised institutions using AI, focusing on governance, risk identification, data quality, testing, explainability and documentation. Pension funds are not supervised by FINMA, but the same principles are a useful benchmark for their own use of AI and for selecting managers.

Exposure to AI is often spread across several asset classes – large technology stocks, venture and growth equity, data-centre infrastructure and power generation. Look-through analysis of underlying holdings helps reveal concentrations that asset-class reporting can conceal.

Digital member portals, automated administration and integrated investment controlling reduce costs and improve transparency for insured members. They also raise requirements for cybersecurity, vendor management and data governance.