In some circles, talk of artificial intelligence means robots coming for our jobs, a dystopian Terminator-esque science fiction vision of the future, and/or a high level concept that has nothing to do with the real world of planning and advice.
And yet, we are starting to see signs of AI slowly but surely finding its way into advice. Platforms and providers are seeing this trend too, and are looking at how they can bring AI into making business more efficient for their organisation, as well as for advisers and clients.
With all of that in mind, and based on adviser feedback, we dedicated a session at our recent #langcatlive Regenerate event to this very topic.
It was a day that saw speakers from the government and the FCA on the same agenda, and where we attempted and succeeded in what may well be an industry first of regulatory lessons from elsewhere held live across three different time zones. (You can watch back any of the sessions you like over on our YouTube channel, and here on our website).
At our recent post-event roundtable, we heard more about how the industry is implementing AI, where it could be deployed in the sector and the controls that need to be in place.
Our panellists were:
Al Ward, head of adviser platform, Aviva
Chris Law, head of UK platform sales, Morningstar Wealth
In our latest State of the Advice Nation report, we asked advice professionals whether they would consider incorporating AI-based technology into their advice process.
Around one in three said they are currently using AI or intend to do so within a year. This increases to over half of respondents who plan to do so in the next five years.
Morningstar’s Chris Law said they are seeing AI adoption at the advice firms they work with, from transcriptions of client meetings to templated suitability reports.
He said “Both personally and from a company point of view, we’re really interested in AI and have started to do a lot of work in this area.
“Without doing platforms a disservice, there hasn’t been that much in the platform industry when I’ve looked at new features and developments that have come out, that’s made me go ‘wow’ in the last couple of years. But I think AI could be that thing. Some of the stuff I’ve started to see us and other platforms look at – it’s really exciting.”
Within Morningstar, the company is investigating the role AI digital assistants could play in responding to questions and providing access to data. For example, an adviser might ask an AI-based system for the ongoing charge figure on a certain portfolio, or request a particular factsheet to be emailed to them.
Chris said that Morningstar has created a digital AI assistant called Mo. This has been done in a “walled garden” environment, using Microsoft Azure’s Open AI service but powered by Morningstar research and data, rather than scraping the whole of the internet, so the company can validate and ensure the accuracy of the information.
Tom Lovett from tech provider FNZ says there is a need for greater education around AI, and for more debate about the practical application of AI in financial services.
Tom said: “There will be people who are very aware of AI-based systems like ChatGPT and applying them in specific use cases. There will also be people who might not necessarily be aware that the technology platform they are using is employing AI as an underlying principle.
“This could mean the AI is doing anything from a seemingly dull automation of processes through to powering more engagement-type tools such as chatbots. We may even see AI in the actual advice process itself as a kind of next-level robo-advice. I don’t think we’ve seen it fully at that level yet, but that could be the direction of the travel.”
Scottish Widows’ Seema Thakkar agreed about the need for education around what AI means for financial advice. She said while it’s exciting that it’s being branded as “the next new thing”, the applications of AI will mean different things to different people.
She said:
“It’s almost like layering it. You perhaps start with automating processes, then it might get more exciting with the use of chatbots, and then actually [you start thinking] what else can it do?
“There’s always going to be a space for a human to be involved in that. It’s important to flag there will be that safeguarding there because we know that a human will need to then check [any outputs], or top and tail. We need to make sure we understand that from a regulatory perspective as well.”
Seema added AI is something Scottish Widows is going to be looking at, whether linking in with tech platforms or “automating the dull stuff”, to Tom’s point, in the background. She says another iteration could be robo-advice – “simplifying products to make them more accessible to customers is absolutely the way we want to go.”
Managing risks and ‘hallucinations’
Our AI session featured LIFT Financial’s Jonny Stubbs, HAW’s Alasdair Walker and Ewan Mitchell, machine learning and GenAI team lead at Aviva. As part of this, Ewan provided an entry point into what we mean by terms such as AI and machine learning.
He talked about the evolution of AI from when a computer makes a decision, to machine learning (essentially “teaching machines to learn”) and where we are now with AI both using powerful computing and the wealth of data available online to build models based on the way our brains function.
Ewan also explained the concept of AI ‘hallucinations’, and why these are something to be wary of when using AI:
“Fundamentally all these models are statistical models. So when you ask ChatGPT to tell you about something, what’s going on under the hood is the model is looking at the data it is trained on and then giving an answer like that. If there isn’t a data point nearby it will extrapolate from the closest it has got.”
He gave the example of getting a very informed answer to a question like: ‘Why does Croatia have such a good football team for such a small country?” and getting the same answer if you ask the same question about Bhutan, even though in that case the answer would be completely wrong.
He added:
“That’s one of the things you’ve really got to watch out for with some of these models. The AI is essentially making it up.”
Al Ward from Aviva picked up on this point during the roundtable discussion, and on understanding the risks of working with AI in the context of advice.
Al said: “Services like Microsoft Co-Pilot, which was mentioned at #langcatlive, are a practical example of the way advisers can protect their data. The question for advisers and planners is: how do you make sure you’re not falling foul of any of the key risks that have been identified? You can then put that human view over everything to ask: does this make sense?
“[Having that process in place] before anything goes anywhere near a client or customer is critical.”
Jonny Stubbs, LIFT Financial and Alasdair Walker, Handford Aitkenhead & Walker debating AI and advice at #langcatlive Regenerate
Quilter’s Jenny Davidson agreed, saying AI is an area where more work is needed across the sector to understand the possibilities, but also the regulatory framework.
She said: “I think we all need to understand the practical applications of AI, but also where the potential issues could occur, and actually where the regulator stands in terms of that as well.
“As an industry this is something we really need to explore, and work with advisers and understand the motivations. So look at why firms would adopt AI and the benefits, but at the same time understand the parameters. Essentially, what are the guard rails you need to stay within? Otherwise [there’s a risk] of creating uncertain outcomes for clients.”
That said, she doesn’t believe fear of the unknown when it comes to AI has to be a deterrent. She made the point that even using AI to create accurate transcriptions of what a client has said in a meeting can have a positive impact, as it allows advice professionals to really understand and capture the meaning behind what’s being said.
Jenny added: “This isn’t about removing the adviser. It’s about creating efficiency to allow an adviser business to do more.”
Chris Law argued that amid the daunting headlines surrounding AI, there were reasons to be optimistic about its use in advice.
He said: “When you read about AI in the press there’s always negativity – jobs will go, people need to be scared about this. And that’s a real shame, because I think that puts some people off. But there’s less talk about the many benefits that AI is going to bring all of us, both personally and professionally.”
Mark Polson is looking forward to the office Christmas party in this week’s #Update whilst also mulling over the flurry of activity from the FCA in recent days.
We provided the research for a report, in conjunction with Parmenion, which reveals how far short of expectations many adviser platforms are falling. The research found that over the last 12 months, 88% of advisers needed to apologise to at least one of their clients on behalf of a platform, and that poor service delivery from platforms impacts 91% of advisers every day.
/ White papers
The Impact of Poor Platform Service
We provided the research for a report, in conjunction with Parmenion, which reveals how far short of expectations many adviser platforms are falling. The research found that over the last 12 months, 88% of advisers needed to apologise to at least one of their clients on behalf of a platform, and that poor service delivery from platforms impacts 91% of advisers every day.
/ White papers
Answering the Call
Service means a lot of things to a lot of different people. It’s so subjective it can be hard to put your finger on. This paper aims to challenge the status quo and inertia that’s built up in the sector for many years.
