Artificial intelligence (AI)—the technology that allows machines to mimic human intelligence and actions to achieve a goal—is at the technological forefront of many industries, from medical research to investment banking.

Although it traces its roots back to the 1940s, it is only during the 21st century that AI left the confines of science fiction to enter the mainstream. Nowadays AI is behind so much of our everyday lives, from predicting the next TV show you’ll watch to understanding the subtle intentions behind your voice-enabled online searches. In fact, projections for AI’s global market could reach $312.4 billion (£236.44 billion) by 2027, despite the impacts of the pandemic.

Although first adopted by businesses in the tech space, AI is also starting to influence the pension industry by helping portfolio managers and administrators oversee and streamline their business processes, while also giving plan participants better tools to manage their pensions. In this article, we’ll examine the potential role of AI in transforming the pension industry.

How Can Machine Learning Transform the Pension Industry

Although AI is still in its early stages in the pension industry, it already shows great potential in streamlining processes and improving customer service, including:

1. Fraud Detection and Prevention

According to government-led task force Project Bloom, pension scams have robbed victims of more than £1 million of their savings in 2019. On average, each victim lost £91,000, but due to the majority of victims not reporting their cases, the agency believes that this number is actually higher.

When the pandemic struck, the financial standing of retirees and almost-retirees became extremely precarious and pension fraud became the third top financial scam, with one in 10 UK adults falling victim or knowing someone who was victimized. Some of the more common pension scams include early pension release (offering to release cash from pensions before 55) and pension review scams (created to persuade a victim to move their money into high-risk and often illegal schemes).

With machine learning tools, these risks can be monitored in real-time, minimizing the likelihood of fraud—a welcome help at a time when the vulnerable are being preyed on. This tech can also identify an individual, authorize payments, or limit access to accounts. Its self-learning abilities allow it to discover any red flags like a breach of regulations, non-compliance, or market abuse.

2. Creating and Delivering Better Retirement Savings Plan

Machine learning works best when it deals with large databases—something that’s becoming more and more available in our industry. For pension plans, the introduction of AI in the process can help them come up with better retirement savings plans from planning to execution:

• Plan Design — Ideally, a retirement plan should meet all its objectives based on a structure that takes into account industry trends, a person’s risk appetite, and their preferred retirement lifestyle. However, analysing this much information is a massive undertaking, which often leads to plan sponsors falling short.

AI can help plan administrators process vast quantities of financial data so they can offer realistic advice and create a plan specifically tailored to a person’s preferences. Tools like this can easily look into the profiles and actions of each participant, making design adjustments as needed.

• Participant Engagement — If a participant won’t engage with their pensions, it’s going to be even more difficult to figure out a plan that will work for them. This calls for better communication strategies.

An AI-powered chatbot can respond better to people beyond the question-and-answer reactive process. Instead, it can provide personalized feedback based on an individual’s updated data, offering available options (e.g., the pros and cons of an investment strategy) that can, in the end, result in making better-informed decisions. When successfully implemented, this can effectively improve participant engagement.

• Plan Governance — Plan sponsors can use AI to improve their business processes so they can fulfil their roles more efficiently. Routine task automation would be possible, especially with monitoring risk, ensuring compliance, and rebalancing trades. AI tools can also be used to better align a sponsor’s products and services based on the participant’s specific needs.

If a participant has a good experience with their plan, they will feel more inclined to stay with that plan even if, for instance, they switch jobs. This will let sponsors keep their participant pool high, allowing them to maintain relatively lower fees.

• Investment Strategy — AI can customize investment strategies and provide more sophisticated valuation methodologies for alternative investments that can be combined with an individual’s pensions (e.g., real estate), instead of simply offering a standard plan. AI can also help sponsors revise existing investments based on a person’s profile.

3. Personalized Dashboard

Through machine learning, AI tools can better understand how an individual interacts with their pension administration dashboards. Think of how popular streaming services are able to predict the next TV show or movie you might like to watch.

Once users log in, an AI-powered dashboard can provide and first show a handful of elements that it thinks are relevant to the person using the pension software, instead of showing them all twenty possible elements that are available. If or when their needs evolve over time, this AI can then adjust and show them relevant elements based on their new preferences.

4. Automating Processes

Automation can be made possible with AI, especially with manual and time-consuming tasks, including:

• Customer service — Chatbots can free up an administrator’s workload as this tool can deal with routine inquiries. And with AI evolving, responses will be more sophisticated, providing personalized feedback based on an individual’s profile.

• Transactions — AI can automate certain processes to automatically trigger certain transactions (e.g., re-underwriting scheme members or reinitiating buyout quotes based on market changes), monitor risk, and make sure the pension stays compliant.

• Analytics — It wasn’t too long ago when an individual wanting to work out their numbers with their advisor would take weeks to accomplish. With AI analytics, however, this process can now be done in real-time, allowing both individuals and administrators to make better and faster decisions. 

With these automated processes, services are ultimately improved which will then lead to better member engagement.

The Challenges of Using AI

While other industries are catching up, there is much reluctance for businesses in the pension industry to take up this new technology due to the following reasons:

• Quality of data — No matter how advanced machine learning is, it is only as good as the quality of data it provides. This is why confidence in data quality is low, as information can be inaccurate or unreliable, which is a source of great worry over a heavily-regulated industry like pensions.

• Data security — Any platform that stores personal and sensitive information will inevitably be put under the security microscope. Apart from worries over potential security breaches, there are also tighter regulations on data (e.g., GDPR) these days which will require businesses to invest in better cyber security protection.

• Longer adjustment period — For algorithms to predict user behaviour, it will need to process huge amounts of data first. Due to this process, it might take several months for AI tools to improve their accuracy and have speedier responses.

• Infrastructure — New technologies bring into question infrastructure compatibility, as organizations with older systems might not have the hardware to accommodate AI software. This means more investment is needed in building a tech setup that can safely house great swathes of data and with enough bandwidth for AI processing.

• Ethics — Big data often goes hand-in-hand with the question of ethics. Some business owners are wary of the potential misuse of data, such as someone who has access to sensitive information acting in bad faith and selling these to third parties.

Current AI Examples in the Pension Industry

Despite the challenges and worries, however, there exist several AI-powered technologies today that are helping organizations, end-clients, and even governments to better handle pension schemes:

1. Alexa

Amazon Echo’s Alexa is not confined to reading out recipes, running apps, or checking the weather anymore. This AI virtual assistant can now help individuals track their pension contributions in the UK.

For those who are saving into Smart Pension or Aviva, users can simply ask Alexa how much their pension is worth today and how much they are paying into it without needing to log into their accounts online.

While the basic functions for both are the same, those in Smart Pension can also make contribution changes via Alexa, while Aviva members can personalize Alexa and check their latest pension value.

2. Japan Government Pension Investment Fund (GPIF)

Japan’s GPIF, the largest retirement fund in the world with $1.5 trillion (£1.13 trillion) in assets, commissioned a study to look into an AI system that would help the government select and assess assets and investment managers. This was made in response to concerns regarding investment performances associated with outside investment managers (i.e., low returns and high fees).

The study’s results showed promise, as the AI system was able to detect and compare investment styles against predicted performances and characteristics in real-time based on a few trading data (e.g., timing, volume, trading items, etc). This could provide a distinct role for AI in short-term trading, which could either replace or enhance an asset manager’s work.

3. Redtail Technology

Redtail Technology, a web-based Client Relationship Management software for the Financial Advisor, recently added AI to its CRM. This new technology has features that can analyse emails, notes, and text messages to better predict client needs.

This new platform analyses sentiment, categorizes key phrases (important pieces of conversation), and pinpoints crucial entities (e.g., brands, products, new products, goals) that are gaining or losing traction with clients. With this tool, advisors can now be proactive and better mitigate any issues that may arise.

AI’s Burgeoning Role in The Pension Industry

While there may be challenges and roadblocks, the long-term benefits of adopting AI definitely outweigh the costs. Although it is in its early years, it will be best for businesses to learn how this technology works and gradually embrace its adoption now, as we expect to see an increasing role of AI in the pension industry in the coming years.