When everyone has AI, being human becomes the competitive advantage
Why more technology should make financial marketing more human – not less
30 September, 2026
TT
16
There is an irony in the current race to personalise marketing with artificial intelligence: the easier it becomes for every company to create personalised content at scale, the less distinctive personalised content becomes.
That is why I believe one of the most important marketing ideas of the AI era is counterintuitive. Over my 20+ year career, I have watched digital marketing move from websites to social, from broad segments to automation, and now from automation to generative AI.
Each wave promised greater personalisation. AI will certainly make personalisation cheaper. It may therefore make genuine human understanding more valuable.
For financial services, where trust and judgement and the consequences of decisions can be significant, that distinction matters.
AI is moving from experiment to infrastructure
The direction of travel is no longer theoretical. The DFSA’s 2025 survey found that 52 per cent of authorised firms in the DIFC were actively using AI, up from 33 per cent a year earlier. Generative-AI adoption rose 166 per cent year on year, and 75 per cent of firms expected to increase AI use over the following three years.
AI is moving rapidly from experiment to infrastructure in the DIFC.
Clients are moving too. EY reported that 71 per cent of GCC investors expected wealth managers to incorporate AI into their offerings.[2] This is not a market resisting technology. It is a market normalising it.
For marketers, the implication is uncomfortable: “we use AI” will soon have roughly the same differentiating power as “we have a website”. I would not build a strategy around simply having the tool. The advantage will come from the judgement behind it – where AI is used, where it is not, and whether the client experience becomes more useful as a result.
Source: DFSA AI Survey 2025
Personalisation is not the same as relevance
Marketing has always wanted to treat clients as individuals. AI gives us extraordinary new tools: propensity models, next-best actions, dynamic content, conversational interfaces, automated summaries and richer segmentation.
But personalisation can easily become cosmetic. We have all seen the temptation: a client’s name in an email, a recommendation triggered by one click, fifty versions of a message because the system can generate them. None of that automatically creates relevance. Sometimes one carefully framed message, built around a real client expectation, is worth more than a hundred technically “personalised” variations.
The test is simple: does the client feel that the institution understands the decision they are trying to make?
That requires context that data alone may not contain. A client who has suddenly increased cash holdings might be nervous about markets, preparing for a property purchase, funding education or planning a business investment.
Behaviour is a signal. It is not a complete story.
Working across multiple global markets over the last two decades has made this especially clear to me. The same observable behaviour can carry very different meanings depending on life stage, culture, family obligations, mobility and financial confidence.
Data can tell us what happened. Good marketing still has to ask why it might have happened – and remain humble about the answer.
| AI makes craft abundant. It does not automatically make meaning abundant. |
A practical model for human-centred AI in financial marketing.
Human judgement becomes a premium layer
The best model, therefore, is not human versus machine. It is machine-enabled humanity.
AI should remove the low-value work that makes financial experiences slow and generic: searching, summarising, routing, drafting, detecting patterns and surfacing relevant information. It should give advisers, service teams and marketers more time to interpret, explain and empathise.
CFA Institute has made a similar point in its discussion of trust in digital wealth management: clients value seamless technology and human empathy, and people continue to place considerable weight on trusted human advice for consequential investment decisions.
That is a useful design principle. Automate repetition. Augment judgement. Personalise context. Bring in a human when stakes, ambiguity or emotion rise.
The new danger is industrial-scale sameness
Generative AI has another consequence marketers should take seriously: it raises the average quality of content while threatening to compress the difference between brands.
If everyone has access to competent writing, images, video, optimisation and personalisation, the market will be flooded with material that is technically polished and strategically forgettable. This is a challenge I think every marketing leader – myself included – needs to guard against. When production becomes cheaper, the instinct is to produce more. The better response is to raise the threshold for what deserves to be produced at all.
Producing more simply because we can is precisely the wrong response. Scarcity moves upstream. Point of view, judgement, proprietary insight, cultural fluency, taste and credibility become more important because they cannot be created simply by increasing content velocity.
In other words, AI makes craft abundant. It does not automatically make meaning abundant.
Trust requires explainability
Financial marketing also has a responsibility that many consumer categories do not. Personalisation is built on data, and AI introduces questions about accuracy, fairness, privacy and accountability.
The DFSA survey found that although 60 per cent of firms had some form of AI governance structure, 21 per cent still lacked clear accountability or oversight mechanisms even where AI use could be critical.[1] For clients, the technical details may be invisible, but the principle should not be: if an AI-enabled interaction informs a meaningful financial decision, the organisation should be able to explain the basis for it and where accountability sits.
Transparency, therefore, is part of the user experience. “Why am I seeing this?” may become one of the most important questions in personalised financial marketing.
The leadership question
The marketing leader’s task is no longer simply to adopt AI tools. It is to decide where machines genuinely improve the client experience and where human judgement creates more value.
My own test for any AI-enabled marketing use case is deliberately simple. Does it make the experience more useful? Does it make the decision clearer? Does it respect the client’s data and expectations? And does it free human beings to do something more valuable? If the answer is yes, AI can deepen relationships. If the answer is simply “it lets us produce more”, we may be optimising the wrong thing.
The coming competitive advantage will not belong to the brands that appear most automated. After years of marketing across different cultures and levels of financial sophistication, I am convinced of the opposite: technology creates the most value when the client notices the understanding, not the machinery.
The best automation makes the experience feel more considered, responsive and, paradoxically, more human.
Humanity needs to be designed, not assumed
There is one more trap. Organisations sometimes assume that adding a human adviser at the end of a digital journey automatically makes the experience human. It does not. A client who has repeated the same information three times, received generic prompts and then been handed to someone with no context has experienced a process failure, not empathy.
Human-centred design means continuity. The system should carry context forward, recognise when uncertainty is rising and give the employee enough information to add judgement rather than ask the client to start again.
The real promise of AI in this context is not simply fewer human touchpoints. It is fewer low-value interactions – so the human moments that remain can be materially better.
The writer is the director and head of Marketing, CEEMEA, Franklin Templeton.
The views and opinions expressed in this article are those of the author and do not necessarily reflect the views of the organisation.
