Pension professionals body sets out five-principle AI framework as sector hits 100% adoption rate
Pension professionals body sets out five-principle AI framework as sector hits 100% adoption rate #
The Society of Pension Professionals (SPP) has published guidance to help pension scheme trustees, advisers, and administrators govern artificial intelligence responsibly, arguing that existing fiduciary duties, rather than new regulation, should drive AI oversight across the industry.
The paper, titled Governance in the Age of AI: A Practical Framework for Responsible Leadership, was released as AI adoption in the pensions sector reaches 100 per cent. According to the SPP’s own annual surveys, 87 per cent of pension firms reported using AI in 2025, rising to 100 per cent in 2026, ahead of any dedicated regulatory regime.
The SPP argues that AI governance should not be treated as a standalone discipline. The framework positions AI oversight as an extension of the fiduciary responsibilities trustees already carry. Embedding controls into existing risk registers and service reviews, the organisation says, is sufficient for trustee boards to pursue AI-driven efficiencies without compromising member protections.
The guidance sets out five core principles for pension schemes integrating AI. The first is proportionality to risk, which calls on schemes to classify AI usage into low, medium, and high-risk tiers; high-risk applications, such as those that directly inform member benefit decisions or financial advice, require enhanced governance and validation. The second is meaningful human oversight, ensuring that automated decision-making processes involve qualified reviewers with genuine authority to challenge or change AI-generated outputs. Third, the framework addresses data security and privacy, warning that confidential scheme or personal data must never be fed into unapproved or public AI tools. Fourth, it urges trustees to update supplier contracts to demand transparency about AI tool usage and human-review policies, and to guard against AI-enabled fraud including deepfakes. Fifth, schemes are asked to issue clear member communications to prevent beneficiaries from acting on inaccurate outputs from public AI chatbots.
Jo Fellowes, Chair of the SPP Administration Committee, said AI had “moved from an emerging technology to an everyday reality across the pensions industry,” adding that the question was no longer whether to use it, but how to do so safely and with appropriate oversight.
The framework follows The Pensions Regulator’s own AI Plan, published in May 2026, which similarly affirmed that trustees remain accountable for decisions and outcomes even when they delegate functions to AI-enabled providers. Both documents reflect an approach to AI governance in the UK pensions sector based on existing duty-of-care obligations rather than purpose-built rules.
Copy link