Artificial intelligence is already bringing about far-reaching changes to the private banking sector, from client relations to internal processes. However, as these technologies become more widespread across the sector, differentiation will depend less on the tools used and more on each firm’s ability to assert its culture, its convictions and its approach to client relations.
Assets under management at Swiss private banks are reaching record levels. Profitability, however, is experiencing ups and downs. It is falling sharply for mid-sized players, and more moderately for large firms and specialist boutiques. Faced with this pressure on margins, artificial intelligence has become one of the key levers that industry leaders are turning to first. But simply adopting it will not be enough, as it will be rolled out in the same way across all firms, whilst the way in which each one codifies and embodies its culture, convictions and rules will remain what sets it apart. AI remains a tool, whilst the firm’s signature is, and will remain, the key element. Two responses are therefore possible: defending the current model, or seizing the opportunity to define the model of tomorrow.
The widespread adoption of AI, coupled with new data architectures and digital tools, is already transforming the wealth management industry through four concrete developments.
Information asymmetry is disappearing. The banker used to know what the client did not know about markets, products and structures. This asymmetry is fading. Clients are equipping themselves with personal AI capable of simulating wealth management scenarios, comparing offers or identifying a solution in the face of an unexpected market movement. Bankers must no longer simply justify their decisions, but must sometimes explain why their approach is more appropriate than that of the client’s personal AI. Wealth managers, whether based at banks or working independently, are gradually equipping their teams to support these client discussions.
Bankers are becoming more efficient and effective. It used to take three days to prepare a half-yearly portfolio review: collecting data, comparing benchmarks and formatting a report. An AI agent can now deliver the same package in a matter of minutes. Efficiency is also a key benefit: an alert sent before a client succumbs to anxiety during a market correction, a message tailored to that specific client. Some interactions are changing in nature. A periodic review conducted by a conversational assistant is more seamless for both the banker and the client than a form to be filled in.
Data is finally flowing freely. Information provided once – risk tolerance, asset classes to exclude, wealth transfer objectives, relationship preferences – no longer needs to be repeated. Data capture becomes automatic, regardless of the channel, and centralised. Bankers, portfolio managers, product specialists, compliance analysts and risk managers: until now, these roles have shared a fragmented view of the same client. Data architectures unify this view in real time, which speeds up decision-making and improves the quality of recommendations.
Internal processes are being transformed from start to finish. From account openings and securities transactions to compliance checks, software development and reporting, agent-based AI speeds up these processes. This is essential, and all players in the sector are already working towards it, meaning it will soon become a prerequisite rather than an advantage.
What these developments have in common is that they free bankers from the day-to-day mechanics of the job, allowing them to refocus on what they have always been but were previously prevented from fully embodying by the operational burden: the trusted orchestrator between the client, the teams and the specialists.
What will make the difference
The entire sector will eventually adopt all of this. Tools alone will not be sufficient to set banks apart. What once distinguished one institution from another – access to information, speed of execution, infrastructure – is becoming standardised. There is a consensus within the sector on the essential nature of human relationships. What must be built, protected or rejected to ensure this does not remain merely a slogan?
Responsibility cannot be delegated. A family trust can now encode its distribution rules in a smart contract: at a given age, a certain share goes to the eldest child. The contract enforces the rule. But when an heir wishes to deviate from it to fund a project the family deems risky, it is up to the trustee to decide – alongside the protector where appropriate – not the system, nor a fully autonomous AI, even when technology makes this possible. The wealth planner does not have this authority either, but it is they who inform this decision, drawing on their understanding of the family’s circumstances. This responsibility to make decisions – and to advise those who do – must remain a human one, by choice.
A firm’s distinctive approach resists standardisation; it is codified without becoming replicable. The unique way in which each individual communicates, interacts with clients, manages affairs and makes decisions in the face of uncertainty can largely be codified within a model. Two firms can provide their advisers with the same market data and produce different recommendations, because one structurally refuses to use leverage beyond a certain threshold, regardless of the opportunity, whilst the other does not. The same logic applies to sector exclusion or a deliberate overweighting, irrespective of the client’s profile. This is not a bias to be corrected. It is a signature to be embraced, a conviction, with the responsibility that comes with it when it costs a performance point, by documenting it and explaining it to the client.
The organisation is becoming flatter. As it stands, the role of junior analyst – who used to compare funds or draft standard reports – is not going to survive. Junior analysts now develop a role drawing on the expertise of senior analysts, guiding the process for the benefit of the whole team, and thereby gaining faster access to senior analysts’ judgement on complex cases. Complementary skills and the gradual codification of expertise are taking precedence over vertical hierarchy – provided that strategic human resources planning is also reviewed, and that the culture and management model evolve, not just the organisational chart.
The experience – for both clients and staff – becomes the key to successful delivery. As technology advances, the client relationship becomes increasingly hybrid. A successful client experience enables a seamless transition between a simple, self-service mode – for certain assets or straightforward transactions – and comprehensive human support as soon as the complexity of the financial, tax, family or inheritance situation becomes apparent. This also requires designing processes centred on the client and the employee. It is better never to roll out a tool that adds a step rather than removing one, and to measure actual adoption rather than simply the roll-out itself. The players who will succeed are those who facilitate this fluid back-and-forth, rather than imposing a one-size-fits-all solution.
Governance is becoming the foundation of trust. A KYC compliance officer can now identify matches that a manual check might miss, such as a namesake of a politically exposed person, even if the first name does not match. But if the logic that subsequently dismisses a flag as a false positive is not auditable, genuine cases may be wrongly dismissed without anyone realising – or only when it is too late. Governance is becoming just as critical as detection itself: the reliability of the models, their explainability, their compliance, and the cybersecurity that protects them.
The time saved must be measured differently. What is measured must change. There is a strong, legitimate temptation to use the time freed up by automation to increase asset or client ratios per manager, and to improve the cost-to-income ratio whilst maintaining staff numbers. But fostering client relationships also means redirecting this time towards active listening – the kind that enhances the quality of advice, builds client loyalty, drives return on assets (RoA) and fosters long-term portfolio growth. The key challenge lies in striking a balance between these two approaches: measuring the time spent on client relationships – beyond mere assets under management – and making this a managerial responsibility, rather than a matter of individual goodwill.
Take control, don’t just go with the flow
Then there is the question of size. Both large and small institutions face the same challenges, but do not have the same strengths. The latter rely on their agility and the absence of legacy infrastructure requiring modernisation – an agility that AI in practice further enhances, giving them access to capabilities they would otherwise have had to build in-house. The former rely on their investment capacity and their ability to scale up. Medium-sized players, for their part, will need to carve out a distinct enough niche so that they no longer have to compete on size. For large institutions, an opportunity is opening up in parallel: that of enhancing their service offering to independent asset managers in the areas of data, governance and AI-as-a-service.
The Swiss financial centre has every interest in seizing this shift as an opportunity to assert a distinctive leadership, rather than simply enduring it. This trajectory cannot be predicted; it must be prepared for. The industry has never been so technology-driven, and yet never so human. The key criterion will no longer be the quantity of data held. It will be the quality of the signature embodied, both humanly and technologically.