9 January 2024

10th ITM Student Essay Competition: [Sponsored by ITM] winner - Scott Hawes

Is Pensions administration a suitable sector for the deployment of artificial intelligence?

The advent of artificial intelligence (AI), propelled by advancements in machine learning and computing capabilities, has breached the confines of traditional tech domains, finding its stride in commercial markets. This technological evolution, while previously enriching sectors like healthcare and tech with predictive diagnostics and speech processing respectively, now beckons new horizons of opportunity for businesses. Particularly in financial enterprises, AI fuses with routine operations to augment capacity, elevate output, and broaden user capabilities, thereby reshaping operational landscapes.

In the pension administration sector, the narrative of AI unfolds with a lens on augmenting administrative efficiency, accuracy, and personalization. The core arenas of pension administration where AI finds its calling include responding to inquiries through chatbots, utilizing enhanced data analytics to review and modify records, employing machine learning for intelligent decision-making in determining entitlement amounts, managing contributions, streamlining the transition of benefits between different pension plans, coordinating pension disbursements, and automating the monitoring of regulatory changes (Blows L. , 2018) (Waite, 2023) (Barker, 2021).

The transformative potential of AI in pension administration is pronounced, emphasizing a saver-centric experience through technological innovations. By embracing AI, the sector is poised to meet modern-day demands, enhance services, and foster member engagement, ensuring resilience and relevance in a rapidly evolving technological landscape (Masters, 2023) (Mantz).

To grasp the evolution of AI in the pensions industry, an examination of documented insights (Blows L. , 2018) is recommended. A GOV.UK report paints a picture of a budding engagement with AI within the sector, with 38% of organizations either planning or piloting AI interventions, 33% yet to embrace the technology, and 27% having fully integrated AI solutions (GOV.UK, 2021).

The sector, primarily driven by private service providers, somewhat trails in AI adoption. Yet, certain initiatives underline a burgeoning interest in leveraging AI:

  1. Predictive Algorithms: The Finnish Centre for Pensions utilized a machine-learning algorithm, achieving a 78% accuracy rate in predicting disability pension retirement, leveraging socioeconomic, earnings, and benefit data (Saarela, 2022).
  2. Automated Administrative Decisions: Pension Danmark has automated approximately 80% of its administrative decisions and aims to increase this percentage.
  3. Chatbots: Keva, a Finnish national pensions company, harnessed chatbots to bolster customer service, aiming for 24/7 support, reducing call volumes, and automating half of the incoming calls and chats, among other goals (Salminen, 2019).
  4. Robo-advisors and Dashboards: The emergence of white-labelled robo-advice systems aids Independent Financial Advisors (IFAs), while AI's role in personalizing dashboards enhances user engagement. For instance, Betterment and SigFig employ AI for tax-efficient transactions and asset allocations, respectively (Koksal, 2020)

PwC’s 2018 Pensions Technology Survey anticipates that over 53% of employers in the sector will channel investments towards automating member communications in the ensuing three years (Blackmore, 2018).

Discussing AI necessitates an acknowledgment of its broad spectrum of applications and interpretations, spanning from machine learning algorithms in predictive analytics to simpler technologies like chatbots (Russell, 2020). The ongoing evolution of the technology, driven by a myriad of user needs, often entails performance trade-offs when fine-tuning AI algorithms for specific tasks (Dacrema, Cremonesi, & Jannach, 2019). This has spurred the creation of specialized AI technologies tailored for distinct applications, moving away from a “one-size-fits-all” paradigm (Caruana, 1997).

The fusion of AI with pensions administration heralds a tide of efficiency and automation, vital for managing the extensive data and computations inherent in pension schemes. As articulated by the financial services group Cardano, AI augments data quality, automates manual undertakings, and births new tools that markedly refine pension administration efficiency. These strides translate to swifter, more accurate outcomes, diminished human error, and the liberation of human resources for strategic endeavours (Mantz). Lane Clark & Peacock LLP (LCP) echoes this sentiment, spotlighting AI's prowess in automating routine chores like addressing member inquiries through chatbots and leveraging machine learning for trend prediction (Waite, 2023).

AI's imprint on pensions extends to personalization, enabling bespoke investment portfolios and enriched member engagement. AltFi accentuates AI's capability to offer pension holders personalized investment portfolios, paving the way for a more tailored pension journey. For instance, machine learning and predictive analytics can guide pension funds to investment avenues resonating with an individual saver's ethos, like eco-conscious companies (Bucksey, 2023). LCP also delves into product tailoring based on data analysis, an effort that significantly bolsters personalization in the sector (Mantz).

AI emerges as a linchpin in elevating risk management within pensions administration by furnishing real-time analysis, thus enabling proactive risk management. Mason Hayes Curran elucidates AI's potential in aiding pension trustees and risk managers with real-time insights into the repercussions of economic adversities like recessions or employer insolvency on pension schemes (Gillick, 2023). Further, discussions on Style of Money and AI-CIO underline AI's role in advancing risk management, cost mitigation, and bolstering efficiency, alongside its impact on evolving liability-driven investing (LDI) -- a cornerstone of risk management in pensions (Veerman G, 2015) (Kerluke, 2023). Lastly, LCP envisions a transformative journey for risk management, member engagement, and client satisfaction through AI, underscoring the long-term vision of fully unlocking AI's potential in the sector (Waite, 2023).

The use of AI systems within pension administration introduces a wide range of challenges and concerns for both service providers, clients, and members. Above all is the cybersecurity risk associated with data breaches, cyber-attacks, and the lack of transparency attached to machine learning.

EY mentions how cybercriminals often attempt to gain unauthorized access to public pension systems through common vulnerabilities, emphasizing the need for enhanced vigilance surrounding pension provider websites, member and employer portals, and staff-conducted investment operations​. (Josef Pilger, 2023) This is essential to preventing unauthorised access could jeopardise the financial security of pension plans and the personal information of the members.

A Deloitte report outlines how AI can augment predictive cyber-intelligence capabilities, including risk-sensing, threat monitoring, and detection. It highlights the need for continuous monitoring and updating to address emerging cybersecurity issues. (Ramachandran, 2019) This allows for service providers to proactively address the cyber risks associated with the integration and development of new technologies.

Additionally, a piece on Pensions & Investments mentions how the SEC adopted new rules requiring enhanced cybersecurity disclosures, which is a nod towards the increasing regulatory compliance requirements in the face of advancing technology​. (Degen, 2023) By introducing new rules, we begin to encourage greater transparency and accountability with our pension administration systems and proactively look to mitigate the short term issues associated with a lack of regulation within the technologies development.

The survey conducted by Mason Hayes & Curran showed a cautious optimism among pension professionals regarding AI. About 72% of the surveyed professionals believed that AI has the potential to enhance outcomes for pension scheme members​. (Gillick & McElligott, Pensions Industry Cautiously Optimistic About AI, 2023) Moreover, the Irish Legal News article further emphasized this cautious optimism, stating that while professionals see the potential in AI, they may not yet be comfortable receiving financial advice from AI systems​ (Gillick, 2023).

The Professional Pensions Administration Survey 2023 revealed rankings for third-party administrators and software providers, where AI could play a pivotal role in enhancing service delivery and operational efficiency. The survey reflects the industry’s acknowledgment of the transformative potential of AI technologies​. (Professional Pensions, 2023).

A report by the Mercer CFA Institute suggests that AI could significantly improve the performance of pensions by reducing operational costs and identifying upcoming risks, which is crucial for maintaining the financial health and sustainability of pension schemes. (Reuters, 2023)

In summation, the integration of Artificial Intelligence (AI) in pension administration emerges as a promising avenue to significantly enhance operational efficiency, precision, personalization, and risk management within the sector. The potential of AI to automate routine tasks, provide real-time analysis, and facilitate proactive risk management underscores its instrumental value in modernizing pension administration, aligning it with contemporary technological advancements. (Mantz) (Gillick, O'Connor, & McElligott, 2023) (Veerman G, 2015)

However, the journey towards fully harnessing AI's potential is not devoid of challenges. The concerns surrounding cybersecurity, data privacy, and the ethical dimensions of AI deployment necessitate a cautious and well-regulated approach to integrating this technology (Josef Pilger, 2023) (Degen, 2023). The industry's cautious optimism, as mirrored in surveys and professional discourse, reflects a conscious acknowledgment of both the transformative potential and the associated challenges of AI (Professional Pensions, 2023) (Gillick, O'Connor, & McElligott, 2023).

Furthermore, the comparatively nascent adoption of AI in pension administration compared to other sectors suggests a measured pace of integration, allowing for a thorough understanding and mitigation of risks (GOV.UK, 2021). The examples of successful AI deployments in pension administration across different regions, as discussed, illuminate the viable pathways and the tangible benefits that can be accrued (Salminen, 2019) (Saarela, 2022).

Engagement with counterarguments and criticisms, alongside an inclusive dialogue among stakeholders, will be crucial in navigating the ethical and operational challenges, ensuring that the deployment of AI in pension administration is conducted responsibly and to the maximal benefit of all stakeholders involved.

As the sector continues to evolve, a collaborative approach among policymakers, pension administrators, and technology providers, backed by robust regulatory frameworks, will be imperative to foster a conducive environment for innovation. This collaborative ethos, coupled with continuous monitoring and assessment of AI applications, will play a pivotal role in ensuring that pension administration not only adapts to the digital era but thrives in it, ultimately serving the overarching goal of securing financial stability for retirees (Ramachandran, 2019).

By embracing AI, pension administration stands at the cusp of a transformative era, where the amalgamation of human expertise and machine intelligence can significantly elevate the quality-of-service delivery, member engagement, and the overall resilience and relevance of the pension sector in the modern technological landscape. (Lumera, 2022) (Masters, 2023)

Notes/Sources

Barker, S. (2021, April 23). Four tech trends to watch in public pensions administration. Retrieved from CIVICA: https://www.civica.com/en-gb/insights/four-tech-trends-2021/

Blackmore, S. (2018, June). The virtuous circle: value for all from pensions technology. Retrieved from The Suite PWC: https://thesuite.pwc.com/insights/pensions-technology-survey-2018

Blows, L. (2018, October). The future is now. Retrieved from Pensions Age: https://www.pensionsage.com/pa/images/PA_Oct_2018_AI.pdf

Blows, P. (2017, September 19). How AI is transforming pensions. Retrieved from reba: https://reba.global/resource/how-ai-is-transforming-pensions.html

Bojarski, M. D. (2016). End to end learning for self-driving cars. arXiv preprint arXiv:1604.07316.

Bucksey, P. (2023, September 23). AI integration: Transforming the sector and empowering savers. Retrieved from altfi: https://www.altfi.com/article/ai-integration-transforming-the-sector-and-empowering-savers

Capita. (n.d.). Taking the hassle out of pensions administration. Retrieved from Capita: https://www.capita.com/expertise/people-solutions/pensions/pensions-administration-and-software/pensions-administration

Caruana, R. (1997). Multitask learning. Machine learning, 28, 41.

Codevilla, F., Müller, M., López, A., Koltun, V., & Dosovitskiy, A. (2018). End-to-end driving via conditional imitation learning. 2018 IEEE international conference on robotics and automation (ICRA) (pp. 4693--4700). IEEE.

Dacrema, M. F., Cremonesi, P., & Jannach, D. (2019). Are We Really Making Much Progress? A Worrying Analysis of Recent Neural Recommendation Approaches. 13th ACM Conference on Recommender Systems, (pp. 101--109).

Degen, C. (2023, 07 26). SEC adopts rule for enhanced cyber disclosure, issues proposal on AI. Retrieved from Pension&Investments: https://www.pionline.com/regulation/ai-cybersecurity-focus-sec-rule-making-meeting#:~:text=SEC%20adopts%20rule%20for%20enhanced,and%20approved%20a%20rule

Development, E. N. (2018, December 12). EUR 29245 - Artificial Intelligence: A European Perspective. Retrieved from Policy Commons: https://policycommons.net/artifacts/2014215/eur-29245-artificial-intelligence/2766658/

Dilmegani, C. (n.d.). 4 Reasons for Artificial Intelligence (AI) Project Failure in 2023. Retrieved from AIMultiple: https://research.aimultiple.com/ai-fail/

Explainable Artificial Intelligence (XAI) Supporting Public Administration Processes – On the Potential of XAI in Tax Audit Processes. (2021). In C. H. Nijat Mehdiyev, Innovation Through Information Systems (Vol. 46). Retrieved from https://link.springer.com/chapter/10.1007/978-3-030-86790-4_28

Gillick, S. (2023, September 19). MHC: Pension professionals treating AI with cautious optimism. Retrieved from Irish Legal: https://www.irishlegal.com/articles/mhc-pension-professionals-treating-ai-with-cautious-optimism#:~:text=The%20business%20law%20firm%20surveyed,outcomes%20for%20pension%20scheme%20members

Gillick, S., & McElligott, B. (2023, September 20). Pensions Industry Cautiously Optimistic About AI. Retrieved from Mason Hayes & Curran: https://www.mhc.ie/latest/news/pensions-industry-cautiously-optimistic-about-ai#:~:text=Pensions%20Industry%20Cautiously%20Optimistic%20About,outcomes%20for%20pension%20scheme%20members

Gillick, S., O'Connor, P., & McElligott, B. (2023, 5 5). I, Trustee: Artificial Intelligence & Pensions. Retrieved from Mason Hayes & Curran: https://www.mhc.ie/latest/insights/i-trustee-artificial-intelligence-pensions#:~:text=,recessions%2C%20inflation%20or%20employer%20insolvency

Gov.uk. (n.d.). Retrieved from https://nationalcareers.service.gov.uk/job-profiles/pensions-administrator#:~:text=Day%2Dto%2Dday%20tasks&text=use%20a%20computer%20system%20to,from%20one%20pension%20to%20another

GOV.UK. (2021, August). Data Foundations and AI Adoption in the UK Private and Third Sectors. Retrieved from GOV.UK: https://www.gov.uk/government/publications/data-foundations-and-ai-adoption-in-the-uk-private-and-third-sectors/data-foundations-and-ai-adoption-in-the-uk-private-and-third-sectors-executive-summary#fnref:3

Gray, J. (2022). Cardano Advisory launches AI covenant assessment tool. Pensionsage. Retrieved 9 21, 2023, from https://www.pensionsage.com/pa/Cardano-Advisory-launches-AI-covenant-assessment-tool-for-smaller-schemes.php

Josef Pilger, H. W. (2023, 07 25). Cybersecurity in pensions. Retrieved 10 22, 2023, from EY: https://www.ey.com/en_us/government-public-sector/us-pension-retirement-and-social-security/cybersecurity-in-pensions#:~:text=Cybercriminals%20will%20often%20attempt%20to,Investment%20operations%20conducted%20by%20staff

Kerluke, M. (2023, 8 1). Why You Should Turn to AI to Boost Your Pension. Retrieved from Style Of Money: https://www.styleofmoney.com/why-you-should-turn-to-ai-to-boost-your-pension/

Koksal, I. (2020, 4). Forbes. Retrieved from How AI Is Expanding The Applications Of Robo Advisory: https://www.forbes.com/sites/ilkerkoksal/2020/04/18/how-ai-is-expanding-the-applications-of-robo-advisory

Lumera. (2022, February 01). Deploying AI for Life and Pensions – the recipe for success. Retrieved from Lumera: https://lumera.com/en/insights/deploying-ai-for-life-and-pensions/

Mantz, F. (n.d.). What do technology advancements mean for pensions? cardano. Retrieved 9 21, 2023, from https://www.cardano.co.uk/perspectives/ai-technology-what-does-it-mean-for-pensions/

Masters, M. (2023, 02 22). Pensions in the blink of an AI. Retrieved from Spence and Partners: https://www.spenceandpartners.co.uk/insights/pensions-in-the-blink-of-an-ai

Nick, B. (2014). Superintelligence: Paths, dangers, strategies. nick2014superintelligence.

Professional Pensions. (2023, May 04). Revealed: The top third-party administrators and software providers in 2023. Retrieved from Professional Pensions: https://www.professionalpensions.com/news/4113307/revealed-party-administrators-software-providers-2023#:~:text=The%20Professional%20Pensions%20Administration%20Survey,had%20experience%20of%20working%20with

Ramachandran, K. (2019, September 11). Cybersecuirty issues in the AI world. Retrieved from Deloitte: https://www2.deloitte.com/us/en/pages/technology-media-and-telecommunications/articles/ai-and-cybersecurity-concerns.html

Reuters. (2023, October 17). AI should cut pensions costs, highlight risks - report. Retrieved from Reuters: https://www.reuters.com/technology/ai-should-cut-pensions-costs-highlight-risks-report-2023-10-17/#:~:text=LONDON%2C%20Oct%2017%20%28Reuters%29%20,Institute%27s%20global%20pensions%20report%20said

Russell, S. J. (2020). Artificial Intelligence: A Modern Approach (4th ed.). Pearson Education, Inc. Retrieved October 2, 2023

Saarela, K. (2022, 8). Work Disability Risk Prediction Using Machine. In A. A. Kevin S, & K. Daimi (Ed.), Proceedings of the ICR'22 International Conference on Innovations in Computing Research (pp. 13-21). Cham: Springer International Publishing. doi:https://doi.org/10.1007/978-3-031-14054-9_2

Salminen, R. (2019, September 17). Partnership in Action: Chatbots for the Pensions Industry. Retrieved from Get Jenny: https://www.getjenny.com/blog/partnership-in-action-chatbots-for-the-pensions-industry

Veerman G, B. G. (2015, November 4). Better Pension Risk Management Through Technology. Retrieved from Chief Investment Officer: https://www.ai-cio.com/thought-leadership/better-pension-risk-management-through-technology/#:~:text=,driven%20investing%20%28LDI

Waite, A. (2023, July 5). AI and pensions: What I learned talking with pensions actuaries. Retrieved from LCP: https://www.lcp.com/our-viewpoint/2023/07/ai-and-pensions-what-i-learned-talking-with-pensions-actuaries#:~:text=The%20technology%27s%20capacity%20to%20automate,value%20than%20either%20working%20alone.

Walsh, D. (2023). Should trustees be worried about the impact of artificial intelligence on master trusts? Pensions Age. Retrieved 9 21, 2023, from https://www.pensionsage.com/pa/images/PA_Sep_23_StandardLifeMtft1.pdf

back to Winning Essays

Last update: 8 January 2024

Scott Hawes
Scott Hawes
Barnett Waddingham LLP
Data Engineer

Benefits Specialist - UK & Ireland

Salary: £40000 - £58000 pa

Location: Hybrid c.2 days per week Berkshire offices

Senior Pensions Administrator

Salary: £35000 - £45000 pa

Location: Greater Manchester, Scotland or Northern Ireland Offices with hybrid working

Senior Associate, Client Servicing - DC Pensions

Salary: £30000 - £35000 pa

Location: Edinburgh

You may also like:

8th ITM Student Essay Competition: runner-up Ieuan Solanki
03 February 2023

8th ITM Student Essay Competition: runner-up Ieuan Solanki

What do you think is the potential benefits and pitfalls of the dashboard in 20 years' time?

Find out More