Unless you’re lucky enough to live an internet-free life, you will have encountered an AI assistant - known as a ‘chatbot’ - at least once in the past few years.

They are designed to please and are usually able to assist with simple tasks, but at times, they can be infuriatingly polite.

They often introduce themselves under a common human name, as if that’ll make me somehow believe they’re not the product of some computer wires buried in a factory somewhere, using up valuable energy resources and destroying the planet.

Personally, they have the capacity to invoke severe rage when I’ve repeatedly asked to speak to a real person, only to be circled back to the same subservient robot, who obliviously asks me what it can help with for the 14th time.

This is just one symptom of the rapid spread of AI across our daily lives. What began as a gimmick around a decade ago has now permeated everything from shopping to sports, even offering its services as a relationship counsellor (I wouldn’t recommend it - it told me I was in the right).

As is often the case, pensions have been slower on the uptake. Only recently has the industry begun to embrace this technology through tools such as digital pension calculators, personalised dashboards, and early attempts at automated financial guidance.

The launch of the UK’s new 'targeted support' regime could take this evolution further, paving the way for AI to play a more central role in delivering financial guidance.

This raises an obvious question: can technology ever truly replace human advice in the pensions and investment space?

Recent evidence suggests not yet.

Just last week, JPMorgan confirmed that it would retire its software-only wealth management brand Nutmeg, with some commentators describing the experiment in 'pure' robo-advice as an abject failure.

Rowley Turton director, Scott Gallacher, said JPMorgan’s move to phase out Nutmeg’s separate identity was “an admission that the pure robo model may not survive in isolation”.

Palantir Financial Planning chartered financial adviser, Eamonn Prendergast, agreed that the human element remains vital.

“Tech alone can’t replace human advice,” he said, adding that the Nutmeg rebrand is “less about one brand failing and more about the limits of pure robo-advice.”

This is not the first time an automated model has faced growing pains.

In both the US and the UK, early robo-advice platforms, such as Betterment, Wealthfront, and Wealthify, built their brands on the promise of low-cost, data-driven investing. Yet most have evolved toward hybrid models, combining automation with human advisers.

The concept of 'robo-advice' also raises regulatory concerns.

The Financial Conduct Authority’s (FCA) reviews of automated investment services have revealed persistent weaknesses in areas such as disclosure, suitability assessments and fee transparency.

Even when the technology works, it can be challenging for consumers to determine whether they are receiving advice, guidance, or merely a set of suggested options.

Transparency and liability are further challenges: if an algorithm delivers poor outcomes, firms must be able to explain precisely why. The regulator has repeatedly warned that “black box” decision-making is unacceptable in the context of consumer advice.

However, the FCA’s recent consultation on easing advice rules aims to allow firms to provide tailored nudges and suggestions to consumers without crossing into fully regulated advice.

While this may open the door to greater automation, it also heightens the need for oversight, explainability and consumer protection.

Don’t get me wrong, I love AI, and there are clear benefits to be gained from introducing more automation into financial advice.

Robo-advice can lower costs, making support accessible to smaller savers who would otherwise be priced out of the market. It can operate around the clock, provide instant responses, and make advice less intimidating, particularly for younger savers who may prefer digital interaction to a face-to-face meeting.

As Gallacher acknowledged, while clients still tend to prefer the reassurance of human advice, robo-platforms may prove a “natural stepping stone” for younger or lower-value investors.

However, pensions advice is rarely straightforward. For the Pensions Age September magazine, I wrote about the under-appreciated role of behavioural science in financial decision-making.

The questions around when to retire, how to draw income, how much risk to take, and how to balance tax, inheritance and inflation concerns are deeply personal and often emotional.

For all its speed and precision, AI still lacks empathy, context and the ability to manage complex human trade-offs.

Consequently, the most likely future is a hybrid model, in which algorithms handle data-driven analysis and forecasting while humans focus on relationships, reassurance and long-term planning.

While I have my concerns about the prospect of robot-led financial advice, I remain optimistic about the potential of new technology to transform the pensions industry.

There is clearly a role for AI in targeted support, in pensions dashboards, and in other digital savings tools that encourage engagement and better outcomes. We just may not be ready for full robo-advice – not yet, at least.

Still, just in case any AI chatbots happen to be reading Pensions Age, please don’t take this the wrong way - you’re doing great, but when it comes to pensions advice, let’s keep the humans in charge for now.

---