Burges Salmon joins Aon to discuss AI in pensions: top takeaways

In July, I joined Aon’s David Burwell, Dom Makemson, and Claire Cochrane to discuss the hot topic of artificial intelligence (AI) governance in pension schemes. The webinar explored The Pensions Regulator’s (TPR) recently published AI Plan, practical use cases for trustees and advisers, and the governance steps schemes can take to manage both the opportunities and risks associated with AI adoption. 

AI tools are becoming increasingly common across the industry and we know trustees are facing important questions around governance, accountability, cyber security and member protection. Below are some of the key themes that emerged during the discussion. 

The Pensions Regulator expects trustees to engage with AI

The discussion highlighted that TPR’s AI Plan, published on 20 May 2026, sets out clear expectations for trustees and scheme managers. In particular, TPR expects trustees to understand where and how AI is being used by or on behalf  of their pension scheme and to establish clear governance and accountability arrangements around its use. 

TPR has also indicated that it intends to monitor AI adoption across the pensions industry, gather evidence on emerging risks and opportunities and continue to develop its guidance in this area. Whilst the UK approach remains principles-based and outcomes-focused, trustees should expect AI governance to become an increasingly important aspect of TPR’s agenda. 

AI presents significant opportunities for schemes

We highlighted the potential application for AI in member engagement. AI-powered tools have the potential to help schemes communicate more effectively with members through innovations such as chatbots, personalised communications and digital assistants. We noted that improving member understanding and engagement remains a key challenge for the industry in approaching its adequacy problems and that AI is likely to play an increasingly important role in addressing that challenge. 

The webinar also explored how trustees and advisers may use AI to support governance activities, analyse complex information, prepare advice and improve operational efficiency. However, the focus should be on using AI to support human decision-making rather than replace it and to ensure that AI use is safe and produces good member outcomes. 

Governance is critical

We emphasised that good governance remains essential in any scheme regardless of how AI is used.  

The panel discussed a practical governance framework built around trustee, adviser and member use of AI. Trustees retain ultimate responsibility for scheme governance and therefore should satisfy themselves that they understand how AI is being deployed with their scheme’s data, including the safeguards and controls in place. Where AI is used to make decisions without meaningful human involvement, schemes may need to consider automated decision-making requirements and the legally-required safeguards. 

Trustee training and the development of an AI usage policy are often sensible starting points. Any governance framework should also be embedded within existing scheme processes, policies and risk management arrangements rather than being treated as a standalone exercise. 

Even apparently straightforward use cases for AI may raise questions around confidentiality obligations, cyber security and data protection compliance. Trustees should be considering the sensitivity of information being processed by an AI tool and the safeguards being applied in each use case. 

Practical actions for trustees

Trustees can take a number of practical steps, including the following: 

• Undertake training to understand AI technologies and associated risks.
• Identify where AI is already being used by the scheme, its advisers and service providers.
• Engage with advisers and administrators about their AI governance arrangements and controls.
• Consider whether an AI usage policy or wider governance framework would be appropriate.
• Review member communications to ensure appropriate warnings are provided regarding the risks of over-reliance on AI tools and AI-enabled scams. 

The interaction between AI and Pensions is a topic we are seeing lots of our clients be very interested in. If you are interested in discussing how AI might impact you, please contact Chris Brown, Partner, Callum Duckmanton, Solicitor, or Ben Jonsmyth, Solicitor, all in the Pensions and Lifetime Savings Team at Burges Salmon. If you are keen to engage in these discussions, we are also preparing an in-person AI event with pensions communication consultancy, Quietroom, on 10 September 2026 in London to explore what is happening and what is around the corner at the intersection of AI in pensions. If you would like to join us, please get in touch with Chris, Callum, Ben or your regular Burges Salmon contact. 

This article was written by Ben Jonsmyth and Chris Brown.