Love it or loathe it, artificial intelligence (AI) can’t be avoided. But to what extent is it impacting the financial services industry and more specifically, the pensions industry?
I’m no tech wizard, but we all have a vested interest in trying to understand AI – and how it can impact our day to day lives (for better or for worse!). To do this, I spoke to:
Albana Ndrecaj: who recently completed the ‘AI: Generative AI, the Cloud and Machine Learning (ML) Ops course at Oxford University and works in the regulatory and financial crime sector.
Simon Brandt: our Head of Technology and key to keeping us at the forefront of technology. It’s no mean feat having the responsibility of cyber and data security on your shoulders!
Duncan Willsher: a Client Director at Vidett who has also recently hosted an internal training session on this topic and has some recent insight into how AI is impacting the pensions industry.
In the first instalment of this two-part series, we discuss different types of AI, how it’s impacting financial services, what it means for the pensions industry and how we’re embracing it to drive efficiencies for our clients at Vidett.
What is AI and what is the distinction between AI, Generative AI and Agentic AI?
Albana: AI refers to computer systems which can simulate varying levels of (human) intelligence. Whilst there is no globally agreed legal definition of AI, for those operating in the UK and EU, the definition in the UK Parliament’s White Paper A pro-innovation approach to AI regulation and the definition in the EU’s Artificial Intelligence Act 2024 (Art. 3(1)), provide a good starting point.
Generative AI (GenAI) is a subset of AI that uses models like large language models (LLMs) to generate new content, such as texts, images, videos etc. GenAI works by identifying and encoding patterns in vast amounts of data, enabling them to generate high-quality content based on the data they were trained on.
One of the rapidly growing areas of AI is ‘Agentic AI’, whereby a GenAI model (or multi-model) is coupled with the robotics to enable it (‘the agent’) to autonomously, or with minimal human intervention, complete certain tasks or actions following its own reasoning and decision making on the best next steps. Think ‘Rosey the Robot’ from The Jetsons!
How does AI already impact financial services?
Albana: In the financial services industry, AI powers customer service bots, fraud detection systems, robo-advisers, and personalised product offerings, with a view to enhancing efficiency and customer experience. Machine learning models are typically used to analyse spending patterns, detect unusual behaviour in real time, assess creditworthiness and help with anti-money laundering processes, and algorithmic trading.
Natural language processing (NLP) enables firms to extract insights from large volumes of unstructured data, such as regulatory texts. AI is evolving rapidly, so it is not unfathomable that your future Compliance Advisory colleague could be an Agentic AI model, to which you probably would not want to feed too much of the office gossip lest it turn up in company reports!
The use of AI is becoming increasingly common and we must remain aware that this pervasiveness presents an increased need for transparency, fairness, and ethical adaptation and use of AI.
We know the pensions industry can move relatively slowly when it comes to embracing change. Where do you see the industry being in 5 and 10 years’ time when it comes to AI?
Duncan: True, but I think the pace of change in tech is so fast now, that pensions can’t help but be dragged along. The differentiator will be who’s embracing it, and who’s simply trying to keep up. AI presents risk and reward – the risks are that it is used to unlawfully gain access to the assets of pension schemes and members, either through fraud or tricking members. I also think the race to adopt AI advice or guidance for members could have bad outcomes. But the rewards are there as well – better decision-making, quicker response times from service providers, and greater personal effectiveness in our work.
So, in my view, we can’t afford to be dragged along, and all need to be understanding its capabilities and embracing it.
To what extent is Vidett embracing AI and how does this help our clients?
Simon: We’re taking a measured approach to adopting AI technologies. We recognise the importance of integrating AI in a way that enhances our operations without compromising security or the value our team brings to our clients.
We have initiated a small-scale AI rollout, focusing on secure AI chat models. This initial phase includes comprehensive training for our staff on the responsible use of AI, ensuring that everyone is well-equipped to use these new tools effectively and ethically.
Our current exploration involves using AI to assist with tasks such as preparing the minutes of meetings and extracting a list of actions from the meeting. The primary goal here is to support our teams, not replace them. By using AI to assist with routine tasks, our team members can focus on providing more value to our clients.
So, there you have it, some insight from industry experts. Join us for part 2 where we’ll cover the opportunities and threats presented by AI, how we make sure our client’s data is secure alongside AI usage and ask the all-important question of: will AI ever replace the value add of a human mind and in particular a pensions expert and/or trustee?
Thank you to Albana, our external contributor for this article. If you’d like to talk to Simon, Duncan or Kat about AI in pensions, why not get in touch?
