AI plan

Published: 20 May 2026

Introduction

Artificial intelligence (AI) technologies have the potential to transform the pensions system and deliver better outcomes for millions of workplace pensions members.

Our vision is that people have a sustainable income in retirement, supported by a pensions system that provides security and value for all. We recognise that as regulators we have a role to play in shaping innovation so that society benefits.

The Pensions Regulator (TPR) will responsibly harness AI technologies to help reduce unnecessary burden, become more efficient and effective, while anticipating and mitigating harms before they occur. We expect the pensions industry to follow a similar approach so that it delivers better value and protection for members.

This plan:

outlines our role and approach to supporting the use of AI in pensions
clarifies our expectations of trustees, administrators and scheme managers
sets out the steps we will take to enable safe AI adoption and AI-powered innovation in the pensions sector
describes how we will harness AI within TPR

We will evolve our approach and update this plan as the pensions landscape changes, and as we learn from our and the pension ecosystem’s experiences.

AI in pensions

Adoption of AI in large segments of the pensions industry is now widespread and accelerating, according to the Society of Pension Professionals’ 2026 AI Survey, which suggests universal use and plans for increased integration into core services.

However, barriers to adoption still remain, notably organisational nervousness, gaps in understanding, cost considerations, and data protection concerns.

Benefits

AI technologies have the potential to radically improve how pensions are administered, how schemes invest on behalf of their members and how members can be engaged and supported to make better decisions. By increasing productivity and enabling better living standards in retirement, they could also contribute to long-term economic growth.

Beneficial uses

Communications and guidance: Schemes and providers are using AI to provide personalised support, for example clearly and accessibly answering members’ questions about retirement options or modelling their outcomes. This can improve decision-making for members who rarely seek traditional advice. Schemes can also use AI to develop targeted and timely communications, tailored to their members’ needs, whether bespoke emails or video statements.

Administration: Schemes and providers are already using machine learning and data analytics to automate routine tasks and improve fraud detection. This trend will continue, leading to faster processing (such as instant transfer calculations or compliance checks) and potentially lower costs. These technologies will also help trustees and providers manage risks more proactively. For instance, by analysing vast datasets to spot funding issues or employer stresses earlier than humans could.

Risks

There are also significant emerging risks associated with AI in pensions. These include:

technology capabilities and adoption outpacing the implementation of robust governance
members using AI for financial planning and advice
increased and more sophisticated cyber-attacks threatening system and operational resilience
AI-generated scams and fraud
bias that could widen existing inequalities

Risks from current AI use

Financial planning and advice: A Lloyds Banking Group survey found that almost 40% of UK adults have used AI in the past 12 months for future financial planning, such as advice on pensions. It also found that 23% of people aged 25-34 trust AI-generated financial advice and insights more than traditional sources.

Unlike authorised financial advisers or providers, general‑purpose AI tools are unregulated. Members’ use of these raises risks around inaccuracy of information, liability for decision-making and how the advice/guidance boundary applies. The long-term nature of pensions means that impacts may take longer to materialise and there can be a compounding effect over a longer timescale.

Fraud: Perpetrators are using AI to amplify existing tactics, such as developing clone websites more efficiently or more convincing phishing. This increases the risk to members’ of disclosing information or authorising fraudulent transactions.

In the future, we anticipate a rise in AI-generated scams. For example, using AI to conduct impersonation fraud, such as setting up fake pension accounts, or create convincing fake personas that approach members with ‘too good to be true’ offers.

Emerging AI technologies, such as AI agents, will generate new use cases for pensions. Adoption by members in everyday life could also create a heightened demand for AI-powered services and products from their pension provider. However, the potential impacts on the pension system of these factors are highly uncertain at this stage.

Our role and approach

We are not an AI regulator. However, it is our responsibility to prepare for, and respond to, the impact AI will have on the pensions system – finding the right balance between facilitating beneficial adoption and innovation, and mitigating AI-related risks. This is necessary to achieve the member and market outcomes we are pursuing through our new corporate strategy.

Our approach to AI

Our focus as a regulator is on supporting the pensions industry to explore and implement AI in a safe, controlled way, wherever they are on their AI journey. How we understand, govern and use AI within TPR informs our regulatory approach to AI in the industry.

We are outcome-focused – AI adoption and innovation must be safe and in the interests of members. We will do this by:

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
harnessing AI to become a more effective and efficient regulator

We are principles-based, appropriate for a fast-moving technology like AI. Guided by principles such as fairness, accountability, transparency and explainability, we can build trust and confidence in AI technologies within the sector.

We are technology-agnostic and evidence-based. As adoption of AI in the pensions industry evolves, we will monitor its use and risks to good member outcomes. We will build a robust evidence base on the use and impact of AI in pensions to inform our regulatory approach towards AI.

We work with our wider pension ecosystem. We will work with government and regulatory partners to ensure consistency, share best practice and emerging risks, and remain responsive to a changing legal and regulatory landscape.

For example, as chair of the Pensions Scams Action Group, we collaborate across government, law enforcement agencies and industry to seek ways to combat pension fraud.

We believe that industry is often best placed to devise common solutions and we will support them in doing so, including through the expert Pensions Data and Digital Industry Working Group we convened in 2025.

What we expect of trustees, administrators and scheme managers

Good governance

Trustees and scheme managers must ensure schemes are well run and deliver good outcomes for members, regardless of which technologies are being used. They remain accountable for decisions and outcomes even when they delegate activities to providers or advisers, such as administrators. They need to understand where and how AI is being used by or on behalf of their pension scheme.

We expect trustees and scheme managers to:

establish clear governance and accountability for the use of AI systems and technologies
assure themselves at all stages 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 as necessary
work to prevent their members being scammed by being aware of AI-driven fraud methods and responding effectively to the evolving fraud threat

We also recommend that trustees, administrators and other 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 where appropriate, to build trust and confidence
stay informed on UK government guidance, emerging standards and cross-industry best practice
share experiences, successes and concerns with others in industry and TPR so that good practice can be adopted and risks managed effectively

Data

We expect trustees and scheme managers to:

have a clear data strategy, allocate resources for improvements, and challenge service providers where standards are not met
ensure that scheme and member data is of high quality, as a critical input into AI-supported processes – see our member data guidance
comply with data protection legislation and guidance, including as it relates to automated decision-making and the use of information in AI systems – see data protection guidance from the Information Commissioner's Office
understand how AI models use and process data and ensuring there are robust controls in place, in line with our cyber security guidance

Innovation

We expect trustees, scheme managers and administrators to seek appropriate and proportional professional advice when considering or implementing innovations.

If you are considering a new business model or commercial opportunity, you should discuss your pensions innovation with us.

We encourage the pensions industry to share their AI successes and concerns with us and each other to increase industry-wide knowledge and capability.

Our workplan

Ensure all schemes are well run and well governed

Good governance is the foundation of safe adoption and use of AI in members’ interests. We will:

work with industry and government to strengthen trusteeship and governance standards, so the sector is prepared for ongoing reforms and any further changes arising from the Department for Work and Pensions recent consultation on trusteeship, governance and administration
publish guidance in 2026 on the responsible adoption of AI for pension schemes, following engagement with the pensions industry on AI use in schemes and their supply chains
work with the Financial Conduct Authority (FCA) to ensure regulatory alignment across the pensions sector and within the pension supply chain, build and maintain joint supervision touchpoints, and share lessons learnt

Put the data building blocks in place for effective AI adoption

AI technologies function by learning from large, high-quality datasets. As set out in our data strategy, we will:

continue engaging with schemes on their data quality controls as they prepare for pensions dashboards and through the Pensions Data and Digital Industry Working Group
apply common and, where possible, open standards for data across the reforms in the Pension Schemes Act 2026
explore making TPR data sets available for industry to access and use

Support and foster responsible innovation

Our role is to ensure that the regulatory environment enables industry to explore new technologies, products and processes that deliver better outcomes for members, without compromising essential protections. We will:

use our innovation service to support those seeking to innovate in pensions, including through the use of AI
work with the FCA’s regulatory and digital sandbox teams on relevant AI use cases, providing specialist knowledge on workplace pensions
ensure that our published content is high quality, machine readable and easily ingested by AI models
work with government and industry to ensure well-designed default regimes, such as the forthcoming guided retirement duty, can help protect members against the risks of biased or flawed AI-powered financial advice or decision-making across their pensions journey

Harnessing AI to become a more efficient and effective regulator

AI is a key enabling tool to help us become a more efficient and effective regulator by augmenting our regulatory activity, flagging and assessing risks, and reducing unnecessary burden.

We are committed to using AI safely, responsibly and ethically, in line with our statutory obligations and aligned with the clear expectations for responsible adoption, transparency and strong governance outlined in the government’s AI Opportunities Action Plan, AI Playbook and Data and AI Ethics Framework. We have established an AI Advisory Council to oversee the ethical use of AI applications within TPR.

Consistent with our digital, data and technology strategy and data strategy, our internal approach to AI covers three areas:

Responsible AI foundations: Establishing and adopting policies and frameworks to ensure the safe, responsible and effective use of AI.
Organisational enablement: Building AI skills and capability across TPR to support confident and responsible use.
Applications of AI: Prioritising and delivering essential, high-impact, and future-focused applications of AI that align with our strategic objectives and regulatory responsibilities.

Case study: Use of AI to detect fraudulent websites

TPR uses an AI-enabled process to help identify and prioritise websites that may be promoting pension scams. This approach has assessed over 2,000 websites, enabling the removal of 29 high-risk sites and reducing manual scanning and triage by around two hours per day.

We will publish our approach to the use of AI within TPR in 2026.

Monitoring and reporting on progress

Monitoring

Understanding how evolving AI technologies are being adopted and impacting the pension system will be critical to ensuring our regulatory approach remains fit-for-purpose and responsive.

We will:

engage with schemes, advisers and providers, as well as international pensions supervisory counterparts, to build a robust evidence base and report on key trends and insights
work with partners such as the Money and Pensions Service and the FCA to understand how different types of members are adopting AI and any risks and opportunities as a result: this includes looking at the equality of outcomes for people with protected characteristics

Reporting

We will report annually on the progress we are making in enabling safe and responsible AI-driven innovation in the pensions industry. This includes:

engagement with our guidance on AI
the number of discussion sessions on AI-enabled innovations held through our innovation service
how we have supported the Pensions Data and Digital Industry Working Group and others on AI
publishing TPR open data, where appropriate

We will also report on any barriers to innovation and AI adoption in workplace pensions, the actions we have taken to address these, and lessons learned. This includes any adjustments to our regulatory approach.

We will develop metrics to measure safe AI adoption and AI-powered innovation across our regulated community for reporting in future years.