The Pensions Regulator (TPR) has published a plan providing guidance on how AI should be used by the pensions industry (the Plan). The Plan also outlines TPR’s role in, and approach to, overseeing the adoption of AI by pension schemes. Further guidance on the responsible adoption of AI is expected to come later in 2026.

Trustees and their service providers may be using AI-driven processes in pensions contexts such as member benefits calculations, transfer requests, or communications, as well as internal and operational uses.

Balancing Opportunity and Risks

TPR aims to shape AI innovation in pensions administration for the benefit of society. It expects AI to increase efficiency and effectiveness, whilst seeking to anticipate and mitigate potential harms before they occur.

Improving Scheme Administration

The Plan highlights how AI can bring potentially significant improvements to pensions administration, investment decision-making and member engagement and support. TPR notes that schemes are already using AI to provide personalised member support, automate routine tasks and enhance fraud protection, and it expects this trend to continue.

Understanding AI Risk

TPR contrasts these potential benefits with the risks inherent in AI, including:

• AI capabilities and adoption outpacing the implementation of robust governance;
• Members using AI for financial planning and advice;
• Increased and more sophisticated cyber attacks;
• AI-generated scams and fraud; and
• Bias that could widen existing inequalities.

In particular, AI tools are mostly unregulated, unlike authorised financial services firms, and TPR cautions that AI can produce inaccurate responses to user queries. Adoption of AI by members in everyday life could also create a heightened demand for AI-powered services and products from their scheme.

TPR’s Approach to AI

TPR sets out its own approach to AI adoption, which it describes as outcome-focused, principles-based, technology-agnostic, and evidence-based. TPR‘s AI workplan will focus on:

• Ensuring all schemes are well-run and well-governed;
• Putting the data building blocks in place for effective AI adoption;
• Supporting and fostering responsible innovation; and
• Harnessing AI to become a more efficient and effective regulator.

Its internal approach to AI will cover three key areas: establishing frameworks and policies to ensure safe and responsible AI use, building skills and capability across TPR to support AI use, and prioritising AI applications that align with TPR’s strategic objectives and regulatory responsibilities.