Artificial intelligence and pensions – how technology is changing the industry

If you have received a member benefits statement on paper through the letterbox, experienced the daunting prospect of dealing with member records that are stored on microfiche or searched through sealed envelopes containing members’ expressions of wish forms, you might think that artificial intelligence is a long way from impacting on the pensions industry.

But whilst some parts of the UK’s pensions industry are reliant on more traditional methods, there are some providers that are exploring and rolling out exciting developments in pensions technology. At the heart of many of these applications is artificial intelligence. 

What do we mean by artificial intelligence?

There are many definitions of artificial intelligence. For the purpose of these Insights, I’m quoting Professors Andreas Kaplan and Michael Haenlein who have stated that artificial intelligence is:

“a system with the ability to do the following:

– to correctly interpret external data;

– to learn from such data; and

– to use those learnings to achieve specific goals and tasks through flexible adaptation.”

Or, in other words, there are a lot of things that can be classed as artificial intelligence other than humanity destroying robots!

Artificial intelligence encompasses:

• machine learning – taking large volumes of data and being able to do things with that data (e.g. anomaly detection – working out patterns in data and finding data that doesn’t fit with the general pattern. In financial services this can help identify fraudulent activity. In pensions, it could help identify over and underpayments and prevent pension scams);
• knowledge representation – teaching systems to do things such as interact with members as frontline member service chatbots. In pensions, this could see a large percentage of member queries dealt with quickly and efficiently without the need for human interaction; and
• optimisation – using data to optimise processes, resulting in efficiency gains and faster service speeds (for example, the way that Amazon optimises delivery routes and how robots in Ocado’s warehouses learn to pick groceries in increasingly efficient ways). In pensions, this could be working out how to streamline processes such as member fund selection or produce the most efficient and effective member communications.

What is artificial intelligence doing today in the pensions industry?

There are plenty of examples of artificial intelligence being applied in the pensions industry. Anyone saving with Smart Pension, Mercer or Aviva can ask Alexa questions about their pensions. Smart Pension members can go even further, using Alexa to make contribution changes.

A number of providers are developing chat bots and conversational avatars to improve member engagement. These could be combined with robo-advisers to produce true engagement on pension savings decisions. Over time, the ability of robo-advisers to learn from past cases and to accumulate ever greater pools of data will improve the quality of their recommendations.

The Department for Work and Pensions is using artificial intelligence to crack down on benefits fraud. It is easy to see how similar developments could be useful in detecting and preventing pension scams.

The sweep of coronavirus has intensified interest in technology, demonstrating how COVID-19 is accelerating trends that were already evident in the pensions industry (see our Insight ‘Will COVID-19 create a new or accelerated normal for pensions?‘ for more on this).

But, despite these examples, it is still the case that the pensions industry in the UK is not at the cutting edge of developments in or the application of artificial intelligence. A legacy of old systems and processes leaves many administrators and members grappling with pen, paper and post. Keen to be part of the technological revolution, some providers are badging mere automated or batch processing systems as AI solutions.

In a survey published by Equiniti in 2019, just half of trustees offered online payslips and only a third of trustees of occupational defined contribution arrangements offered their active members online switching or self-service.

One of the main reasons for this is the quality of data. Only half of the survey respondents with defined benefit arrangements knew their common data score, dropping to a mere 18 per cent for defined contributions schemes.

Towards a pensions data revolution – what are the next steps?

Pension dashboards will be the legislative nudge that the pensions industry needs to improve data quality and embrace the next stage of technological innovation. Pension dashboards are in the beginning phases of a long road to delivery across the industry. But, with legislation in the Pension Schemes Bill 2019 – 21 and an active delivery programme, they will be part of the pensions industry in the medium term.

The Pensions Dashboards Programme (part of the Money and Pensions Service) is responsible for:

• the technical architecture; and
• standards and services based on the needs of users.

The Pensions Dashboards Programme is also responsible for delivering:

• the Pension Finder Service;
• the Identity Service (i.e. the service that will enable individuals to prove who they are in order to ensure that the right people have access to the right information); and
• governance registers.

The Money and Pensions Service will be responsible for building a pension dashboard based on the digital architecture set up by the Pensions Dashboards Programme.  

All of this is expected to be delivered over the next four years, meaning that pensions in the UK will have open data by the middle of the decade.

Once the industry has high quality, open, standardised, accurate and digital data, the possibilities for transformation and disruption are enormous. If the risks are properly managed, the result will be an industry that provides much more choice, higher levels of engagement and customer satisfaction and, hopefully, the increased savings that are needed for people who are living longer.