AI and the IFA: What could ChatGPT mean for financial advice?

It’s the topic of the moment and a breakthrough in technology that (some say) is set to be as seismic as the creation of the internet. 

Generative artificial intelligence (AI) was launched on the world barely six months ago. But its potential to transform many industries is already being widely predicted.

So, what could it mean for financial advice? 

Could it take on certain tasks and enable advisers to focus their human resources on more value-added activities? How are advice firms planning to use it (if at all)? And what do we at M&G Wealth think of its potential? 

Over the coming months, we explore all these questions. Let’s start here with the basics: what exactly is it?

What is generative AI?

‘Traditional’ artificial intelligence (AI) is programmed with rules in order to perform specific tasks, such as running a web search engine to driving a driverless car or playing chess. 

Generative artificial intelligence (generative AI) goes further. Rather than just performing pre-defined tasks, it is capable of creating a variety of new content and data in seconds by harvesting existing data sets – from text to images to music. 

Using deep machine learning, generative AI’s particular appeal comes from its ability to deliver highly human-sounding, creative, insightful and empathetic responses. And the rate at which it appears to be getting smarter is breath-taking.  

The key players

ChatGPT – a natural language tool driven by AI technology - is currently leading the market. Launched as a prototype by the Microsoft-backed OpenAI only last November, it reached 100 million users in two months. Its latest version, GPT-4, is already reported to be capable of beating 90% of lawyers trying to pass the US bar exam.  Read the full article here. 

The technology is spreading quickly. Microsoft is embedding Copilot, an AI-powered digital assistant, into everyday apps such as Word, Excel and PowerPoint. In May, Google announced that its ChatGPT-like AI assistant, Bard, was now available in 180 countries with the ability to export its responses directly to Gmail or Google Docs. 

Opportunities for advice firms

Potential applications for the financial advice sector are quickly becoming evident. Repetitive writing tasks are the obvious target, from building client suitability reports to generating regular market commentary. 

Generative AI’s ability to sound warm and human may also be a powerful acquisition tool. In April, Citywire reported on a firm offering to use ChatGPT as a lead generation tool to screen new clients. Read the full article here. 

Delivering the human touch

However, right now, generative AI looks unlikely to replace human-led financial advice, especially given robo-advice’s decade-long struggle to gain a market foothold. 

Generative AI may be able to help generate client reporting, using a firm’s own centralised investment proposition as its core data set for example. But as we reported in our Consumer Duty Guide to Fair Value soft skills, direct knowledge of clients’ circumstances, and face-to-face interaction are what clients often value most in their advisory relationships. 

As yet, artificial intelligence cannot outgun emotional intelligence.

Human oversight is also still critical, given generative AI’s continued tendency to generate highly plausible but factually inaccurate responses (known as ‘hallucinations’). Privacy concerns and security threats are also an issue. 

Recognising the potential

There is a natural tendency to overestimate the uptake of new technologies (remember the Internet of Things and our fridges ordering our groceries for us?). But it’s also dangerous to underestimate the impact that AI is already having on the world of work. 

BT recently announced plans to cut its workforce by as much as 55,000 by 2030, with about 10,000 jobs being replaced by artificial intelligence. Which comparably sized financial players are likely to follow suit?

Generative AI is one piece of tech that advice firms certainly need to monitor (and experiment with) to see where it might build in benefits and efficiencies. We’ll explore its potential applications for firms in future articles.