Artificial intelligence has quickly become part of everyday life. Now used in almost all areas of business—drafting emails, summarising documents, explaining technical concepts, recording meetings—AI is changing the way we work. It is also changing the way we learn.
It isn’t surprising that candidates studying for professional qualifications are embracing these tools. Used well, they can explain difficult concepts in plain English, create revision plans, generate practice questions and even test understanding (though, usually unreliably). For busy professionals trying to fit study around work and family life, that’s an incredibly attractive proposition.
The question, however, isn’t whether candidates should use AI.
The question is whether we’re beginning to mistake finding knowledge for owning it.
Professional exams have always required candidates to demonstrate technical understanding. Increasingly, though, we’re seeing a generation of learners who can access information almost instantly. Need to know the Annual Allowance? Ask AI. Forgotten how tapering works? Search for it. Unsure how Business Relief operates? A chatbot can produce an explanation in seconds.
On the face of it, this sounds like progress but it raises an interesting question.
The question is whether we’re beginning to mistake finding knowledge for owning it
Once you overcome the problems of hallucinated information, out-of-date rates and allowances, or answers relating to foreign legislation: if candidates never have to wrestle with the material themselves, are they actually learning it?
Psychologists refer to something called the “illusion of competence”. It occurs when information feels familiar because we’ve recently seen or read it, leading us to believe we understand it more deeply than we really do. Reading an AI-generated explanation can create exactly this effect. The content is often clear, well-structured and convincing. It feels as though learning has taken place.
Yet when candidates sit down in an exam without access to AI, many discover that recognising an answer and recalling it independently are two very different skills.
Professional examinations are not designed to test whether candidates know where to find information. They assess whether candidates can retrieve relevant knowledge from memory, interpret the question, and then apply that knowledge appropriately under pressure.
Those are cognitive skills that still require practice.
This is particularly true as candidates progress towards more advanced qualifications. At this stage, success depends less on remembering isolated facts and more on understanding how technical concepts interact.
An adviser may need to consider taxation, pensions, trusts and investment planning simultaneously within a single client recommendation. AI can certainly explain each of those topics individually. Whether the learner has developed the judgement to bring them together is another matter entirely.
That distinction becomes even more important once candidates enter practice because clients rarely ask textbook questions. They present complex, often incomplete situations requiring interpretation, prioritisation and professional judgement. The value of an adviser has never been simply knowing the legislation. It lies in understanding how that legislation applies to the individual sitting opposite them. AI cannot develop that judgement on someone’s behalf.
This is not an argument against using technology. Quite the opposite.
Future advisers will almost certainly need to become highly proficient users of AI. Firms are already exploring how artificial intelligence can support research, summarise information, improve productivity and enhance client communications. Some are even using AI to write suitability reports and manage client communications. Some of these tools will undoubtedly become part of everyday professional life.
However, the role of AI should be to support thinking, not replace it.
There is an important difference between asking AI to explain a difficult concept and asking it to answer a practice question for you. The first supports learning. The second risks bypassing the very mental effort that creates learning in the first place.
Educational research has consistently demonstrated that retrieval practice—forcing ourselves to recall information without prompts—is one of the most effective ways to strengthen long-term memory. The struggle involved in remembering isn’t a sign that learning has failed; it’s the process through which learning occurs.
Candidates can now receive instant answers without ever experiencing the productive difficulty that helps knowledge become embedded. While this makes studying feel easier, easier does not necessarily mean more effective.
There is another consideration too.
Generative AI is impressive, but it is not infallible. Financial services is built on legislation, regulation and taxation that evolve continually. AI systems can present outdated information or generate entirely plausible answers that are simply wrong. Perhaps more concerning is that they often do so with remarkable confidence.
Candidates without sufficient technical understanding may not recognise those inaccuracies.
Ironically, the people most likely to benefit from AI are often those who already possess a solid technical foundation, because they have the knowledge needed to challenge, verify and interpret its outputs.
Rather than asking whether AI belongs in learning, we should be asking how candidates can use it responsibly.
Can it help explain a concept? Absolutely.
Can it generate revision questions? Certainly.
Can it challenge understanding or identify weaker areas? Without doubt.
But should it become a substitute for building technical knowledge, practising retrieval or developing professional judgement? Probably not.
Professional qualifications have never simply been about passing an exam. They are about laying the foundations for a career in which clients place enormous trust in the advice they receive.
Technology will continue to change how we learn, just as it is changing how we work.
The challenge for candidates is ensuring that convenience doesn’t come at the expense of understanding. Because ultimately, the best advisers won’t be those who know how to ask AI the right question.
