How can artificial intelligence help pension schemes?

Whilst artificial intelligence (AI) was first adopted by technology businesses, statistics show that professionals are optimistic about the impact it can have in transforming the pensions industry.

In response to a survey conducted by Professional Pensions, 17% of respondents believe that AI will revolutionise the pensions industry a great deal, while a significant majority of 60% believe that AI will revolutionise the industry to some extent.

AI has the potential to help portfolio managers and administrators with risk management, client satisfaction and automating processes, along with increasing member engagement, which is often cited as a persistent and frustrating challenge to the industry.  According to research by Atlas Master Trust, 72% of respondents do not actively engage with their workplace pension. A further White Paper found that 28% of respondents never review their pension and try not to think about it at all. AI can offer a more engaging potential solution to this problem, by providing members with a personalised approach to communication, encouraging them to interact with their pension provider.

In this article, we will look further into how AI can help increase member engagement and tackle this industry-wide problem.

Increasing member engagement

There are four key ways in which AI can be harnessed to help encourage participant engagement: plan design, participant engagement tools, plan governance and investment strategy.

In terms of plan design, generative AI works best when it deals with large databases of information. Retirement plans should consider factors such as industry trends, preferred retirement lifestyle of the individual and their risk appetite. With access to an individual's financial data, AI can help administrators create a plan specifically tailored to a person's preference. AI can help capture and analyse information and track plan design weaknesses, which can translate to better serving a participant's long-term interests, driving up engagement.

When looking at specific participant engagement tools, pension communications are often poorly targeted, dissuading members from interacting with their provider. AI could be used to deliver personalised communications to individuals, encouraging them to become more proactive and engaged. 

AI could also be used to provide an AI-powered chatbot. This could provide tailored feedback to individuals and evolve from a reactive service, such as dealing with a person who needs assistance with a query, to being a proactive service which is updated on participant milestone such as salary raises and can offer informed advice.

Plan sponsors can harness AI to ensure they fulfil their roles efficiently and improve business processes, which in turn leads to strong plan governance. Administrative tasks such as monitoring risk and ensuring compliance could be automated and will ensure that sponsors are able to fully execute their fiduciary obligations and that participants have a positive experience with their pension providers.

Lastly, AI can be leveraged to help individuals decide on an investment strategy. AI can customise strategies and provide more sophisticated valuation methods instead of simply offering a standard plan. For instance, AI can account for alternative asset classes such as real estate to truly create a plan based on an individual's profile and expand diversification options for participants.

Challenges

All of the above ways show how AI can be used to drive up member engagement. However, there are still barriers to rolling out AI in the pensions space.

One barrier would be the availability of high-quality, accurate digitalised data. This could be a challenge for many schemes and administrators, as often pension schemes rely on legacy systems and data. There will likely need to be huge investments in systems and processes in order to ensure data is used effectively to deliver results for participants.

A further risk of using AI in the pensions sector is data protection and cyber security. Pension schemes hold a great deal of personal and sensitive data about members. Research indicates that UK pension schemes have seen the biggest increase in cyber security breaches, reporting a 4,000% rise in data breach reports to the Information Commissioner’s Office, going up from six in 2021/22 to 246 in 2022/23 (to read more about cyber security and pension schemes, read our previous blog here).

Conclusion

Despite these potential challenges, the long-term benefits of adopting AI have the potential to re-shape the industry.  As Maurice Speer, Mercer pension risk transfer principal says: "AI could help improve the member engagement problem in pensions… but with creative personalisation of communications for individual circumstances and preferences, chatbots for immediate access to routine queries, and powerful predictive analysis to anticipate proactive communication and support, AI has the potential to transform member communications in the pensions industry".