AI and UK Pension Schemes: Key Actions for Trustees

The use of Artificial Intelligence (AI) is accelerating within all sectors. Whether its driverless cars, autonomous farm machinery (as seen on the latest series of Clarkson’s Farm) or the increasing use of generative AI tools, there is no doubt that the AI revolution is well underway.

According to the Society of Pension Professionals’ 2026 AI survey, adoption of AI in large segments of the pensions industry is now widespread and accelerating. In particular, AI can help to improve efficiency and provide more personalised support for members.

In response to this trend, the Pensions Regulator (TPR) has recently signalled its approach to the use of AI by pension schemes, in its AI plan. It recognises that

“AI technologies have the potential to transform the pensions system and deliver better outcomes for millions of workplace pensions members”. 

However, as well as seeing the potential benefits, TPR is also clear on the risks. It places the onus on trustees to identify and understand how AI is being used in connection with their scheme and to put in place effective controls and a robust governance framework.

So, what does this mean for pension scheme trustees? 

What is TPR’s AI Plan?

In its AI Plan, TPR sets out its current thinking and approach on the use of AI by pension schemes. In particular, it:

• highlights some of the potential benefits and also the risks associated with the use of AI
• outlines TPR’s role and approach to supporting the use of AI in pensions
• clarifies its expectations of trustees, administrators and scheme managers
• sets out the steps it will take to enable safe AI adoption and AI-powered innovation in the pensions sector, and
• describes in high-level terms how it will harness AI itself.

The Plan makes clear that accountability for the use of AI and related outcomes remains with trustees, even where AI tools and solutions are being used by service providers, such as administrators, or advisers.

TPR has signalled it plans to engage with the industry over the summer to understand more about how AI is being used. It then plans to issue further guidance later in the year. Trustees however should not wait for this – there are actions they should be taking now. 

What are trustees expected to do?

TPR expects trustees to: 

• understand where and how AI is being used by, or on behalf of, their scheme
• establish clear governance and accountability for the use of AI systems and technologies
• assure themselves that their administrators, service providers and advisers have similarly robust governance arrangements in place
• carry out rigorous testing, assurance and ongoing monitoring, both at the point of implementation and on a regular basis afterwards
• identify and evaluate risks, make sure appropriate controls are in place, review these regularly and adapt them as necessary, and
• work to prevent their members being scammed, by being aware of AI-driven fraud methods and responding effectively to the evolving fraud threat. 

TPR also recommends that trustees, administrators and other service providers:

• invest appropriate time and resources to understand AI technologies, including responsible uses, limitations and risks  
• are transparent with scheme members and stakeholders about AI use by or on behalf of their scheme, where appropriate, to build trust and confidence
• stay informed on UK government guidance, emerging standards and cross-industry best practice, and
• share experiences, successes and concerns with others in the industry and TPR so that good practice can be adopted and risks managed effectively.

TPR recognises that the quality of the outputs from AI is dependent on the quality of the data that is put in. Therefore, it highlights the need fortrustees to:

• have a clear data strategy, allocate resources for improvements, and challenge service providers where standards are not met, and
• ensure that scheme and member data is of high quality.

It also reminds trustees of the need to:

• understand how AI models use and process data and ensure there are robust controls in place, in line with TPR’s cyber security guidance, and
• comply with data protection laws and guidance, including as it relates to automated decision-making and the use of information and data in AI systems.

How should trustees apply this in practice?

We suggest trustees start their AI governance projects by taking reasonable steps to:

• understand how AI works, how it is developing, the potential benefits for scheme administration and member outcomes, and the associated risks and regulatory expectations – this is likely to involve training for the trustees, and
• understand how AI is currently being used by their scheme – this could include sending out a questionnaire to their scheme’s service providers and advisers to understand how they are using AI on work they do for the scheme and the safeguards and internal governance they have in place.

Trustees should then:

• put in place a governance framework which works for their scheme and how AI is used now (and which is capable of evolving as the technologies and their uses develop) - this includes putting in place an AI policy
• update their agreements with their scheme administrator, advisers and other service providers to cover the use of AI, including the need for approvals, governance and controls
• explore with their investment advisers how the development of AI impacts scheme investments (both at a manager level, and the potential impact on risk adjusted returns of investments over different time horizons)
• where schemes have scheduled cyber reviews of advisers and service providers, build AI into those, and
• strengthen the quality of scheme and member data used by AI, where necessary. 

Looking ahead, trustees will need to consider how they will monitor the evolving use of AI by their scheme and by their service providers and advisers, so that they can harness the opportunities this creates while continuing to effectively manage the risks.

They will also need to decide how to inform members about how AI is being used to ensure they continue to meet their transparency obligations.

How can we help?

We have real expertise and strength in depth across our pensions, commercial, data protection and consulting teams on the opportunities for pension schemes with their use of AI and how to manage the risks. Please speak to your Eversheds Sutherland adviser to discuss how we can help you put in place effective, tailored and proportionate measures for your scheme or contact one of our specialists.