Clients over 60 and digital discomfort: artificial intelligence is changing the advisor’s role

An entrepreneur may have built a company, negotiated financing and managed significant assets, and yet find themselves struggling with a banking procedure on a phone. They know how to send a message, check email and join a video call, but they are not equally sure how to tell a document to be downloaded from one that is to be signed, or a simple notification from a request that authorizes an operation. This is a plausible situation even among educated people with strong financial expertise. Technological discomfort often stems from the gap between what one is used to doing and what the service demands, without leaving time to understand. When that gap concerns money, the operational difficulty also becomes a problem of asset autonomy.

Age helps explain the phenomenon, but it does not allow us to describe all sixty- or seventy-year-olds in the same way. According to Istat, in 2025 72.5% of people aged 65 to 74 used the Internet in the previous three months; among those over 75, the figure was 35.7%. The report also measures skills beyond simple connection: understanding information and data, communication, content creation, security and problem solving. This is an essential distinction for understanding digital discomfort. Being connected does not, by itself, prove that someone can solve a problem, protect data or assess information; and not knowing an application does not erase the professional skills acquired over a lifetime.

Familiarity with the phone is not the same as digital literacy

WhatsApp and email meet very recognizable needs: talking to someone, receiving a message, sending a photo. Repetition makes these actions familiar, but it does not automatically transfer the same confidence to other environments. A person may use a phone every day and still not know where a contract was saved, which version was edited, or how to regain access to a service after changing devices. The problem emerges when the steps are linked together: opening an attachment, checking where it came from, saving it, comparing it with another document and possibly authorizing a choice. Digital literacy also means understanding this sequence, recognizing which actions can be reversed and which have economic or legal consequences.

The answer cannot be to ask everyone to fully master Word or Excel. Knowing every function of a program is not a reasonable measure of a client’s autonomy. For someone who needs to monitor a wealth statement, what matters is understanding which data have been entered, distinguishing an amount from a formula and noticing if the result does not add up. Similarly, a document that is neatly laid out may contain incorrect information or an incomprehensible clause. The gap is therefore also cultural: it concerns the way we question tools, read responses and ask for explanations. It can affect a young person accustomed to social media as much as an older professional, even if it appears with different frequency and different levels of difficulty.

When the service changes, the client may lose confidence

Discomfort increases when the interface changes precisely while someone is learning to use it. A code arrives on a different channel, a confirmation moves, a procedure is replaced and the instructions seem to assume that everything is intuitive. For those who do not have a map of how the system works overall, each change can feel like a new obstacle. The fear of making mistakes then leads people to put things off, give up or ask someone else to do everything. Shame and embarrassment can also weigh heavily: someone who has always had major financial responsibilities may find it hard to admit they do not understand an elementary step. Service quality is also measured by whether people can ask for help without being treated as a problem to dispose of.

The progressive digitization of financial relationships can thus shift activities onto the client that were previously carried out with an operator. Some people gain time and freedom; others encounter a new dependence, especially if the alternative channel is hard to reach. It is therefore not enough to measure how many accesses have been moved to an app: it is necessary to verify whether the client can complete the procedure and understand what they have authorized. A successful login may be the result of an occasional family member stepping in, without the account holder having gained autonomy. For an intermediary, accompanying this transition requires clear instructions, accessible assistance and operating methods consistent with the person’s actual abilities.

Artificial intelligence can lower the barrier of commands

The arrival of artificial intelligence introduces a different possibility from the traditional need to learn a program. In conversational systems, a problem can be described in everyday language, an explanation can be repeated and one can proceed by trial and error without knowing the path through the menus in advance. For some people this may be less demanding than memorizing formulas, shortcuts or command sequences. The same question can be rephrased until the result becomes understandable. This does not mean the experience is automatically easy for everyone: device availability, language, reading ability and trust in the tool still matter. But the possibility of starting from a concrete need rather than the software’s workings opens up a significant space for inclusion.

For example, one can ask for an explanation of the general meaning of the items in an account statement, using a sample document with no personal data, or for an explanation of how a change in costs affects a hypothetical balance sheet. A spreadsheet can be tackled by asking which information is needed, how to organize it and which formula to use for comparison. In these cases, the tool can help even someone with limited technical skills build an initial structure around the problem. The advantage is not to skip all learning, but to make it closer to the person’s own language. It is still necessary to check the steps: a suggested formula should be tested on known values, an explanation should match the document and a calculation should make the assumptions used visible.

The difficulty shifts to the ability to judge the answers

A simple interface can hide a complex task. AI produces tidy, convincing answers, but the clarity of the presentation does not guarantee that figures, legal references and conclusions are correct. Someone who has always felt excluded from technology may feel relieved to finally find an interlocutor who responds patiently; that very availability can lead them to trust it more than is reasonable. The new skill lies in framing a request precisely enough, understanding what information is missing and recognizing when the result needs external verification. Asking for a source and a date helps, but even a well-presented reference may be inaccurate: the check must go back to the original document, not stop at the citation produced by the system.

In investment matters, this distinction has concrete consequences. In its March 2025 warning, ESMA (European Securities and Markets Authority, European Securities and Markets Authority) warns that publicly accessible AI tools can produce indications based on incomplete, incorrect or outdated information. An answer that seems personalized is not the same as the service provided by an authorized intermediary, with the related responsibilities and safeguards. For the client, asking for an explanation of a financial concept and entrusting a general-purpose assistant with investment selection are very different decisions. The advisor should make this boundary understandable without demonizing the technology: using it to prepare questions can improve the meeting, while treating it as a guarantee of returns exposes one to a dangerous misunderstanding.

The advisor must support without taking the client’s place

Professional value emerges when assistance gradually increases a person’s ability to understand. Solving every difficulty by taking the client’s phone and completing the procedure on their behalf may seem efficient, but it leaves dependence untouched. It is more useful to explain the purpose of the operation, identify where the person gets stuck and have them repeat the step in safe conditions. Asking them to describe in their own words what they are about to authorize also helps reveal misunderstandings that a signature does not show. The goal is not to turn the advisor into an IT technician: it is to prevent an operational obstacle from undermining the meaning of the wealth decision and to avoid mistaking haste for informed consent.

Financial competence and digital competence must be observed separately. An experienced entrepreneur may understand investment risk and still need support accessing documents; a person who is very quick on a smartphone may not understand costs, illiquidity or the possibility of loss. The information collected in the relationship should accurately reflect both situations. The support needed should be agreed in advance, without attributing incapacity on the basis of age and without inferring financial knowledge from technical ease. Written explanations, a recognizable contact channel and instructions tied to the operations actually used are often more useful than a full demonstration of all available functions. The measure of success is what the client can then do and understand on their own.

Help must not become a handover of credentials

The need for assistance does not justify sharing passwords or the codes that authorize operations. Asking a child, colleague or professional to solve a problem may be legitimate, but the powers to operate on the account must pass through the formal tools provided by the intermediary. Banca d’Italia includes the protection of credentials among the behaviors relevant to digital finance. In an entrepreneurial family, the distinction is particularly important: the same person may informally help a parent with email and then find themselves, without any clear agreement, handling bank communications and confidential documents. Defining who can see, prepare or authorize each operation protects everyone and reduces conflict.

Caution also applies to the data entered into AI assistants. Identity documents, full account details, credentials and detailed information on family assets should not be uploaded to a general-purpose service merely because the conversation feels private. Protection depends on the service, the contractual terms and the actual settings, not on the reassuring tone of the response. To learn, one can use fictitious examples; in professional activities, one needs tools and procedures authorized by the organization, with checks on the data processed and who can access it. The professional must be able to explain why a request is appropriate in one environment and not in another. Digital education also includes the ability to forgo a shortcut when it requires exposing information that is not necessary.

Recognizing a voice is no longer enough to recognize a person

The new ability to imitate voices and images makes some of the signals on which everyday trust was based less reliable. The fact sheet on financial fraud in the AI era, issued by the European authorities and published by Banca d’Italia, describes messages, videos and calls that can mimic an advisor or a family member. Protection must therefore rely on verifiable procedures: if an unexpected request for money or codes arrives, break off the contact and call back using a number already known, obtained independently from the message received. A familiar voice, the correct logo or impeccable grammar are not enough. A well-organized professional relationship can establish in advance how requests will be communicated and which requests will never be made.

Here too, the idea that vulnerability coincides with age should be avoided. Haste, concern for a family member, the expectation of a gain or trust in an interlocutor can affect people of different generations. Digital discomfort can make it harder to respond to an unusual request, but excessive confidence can produce the same outcome. Education should therefore train people to stop, check and ask for a second opinion, rather than merely memorizing the look of the latest scam. If the client knows whom to turn to without feeling judged, they are more likely to report a doubt before carrying out the operation. This concrete availability is also part of the quality of the relationship.

Practical training and professional responsibility must advance together

Not all support should fall on the individual advisor. In June 2026, Banca d’Italia and the Department for Digital Transformation launched a collaboration on digital financial literacy, with training for facilitators at the Punti Digitale Facile. The initiative covers secure access, banking operations, document reading and fraud prevention. It is a useful reference because it connects learning to real tasks, rather than to the mere availability of connections and devices. A professional can know the services available in the area and direct the client to appropriate support, while retaining their own responsibility for wealth decisions. Collaboration between different skills prevents a basic difficulty from remaining unresolved or being handled through indiscriminate informal delegation.

The advisor must also know the tools they propose or use. ESMA’s May 2024 guidelines on AI in investment services emphasize the responsibility of intermediaries and the primacy of the client’s interest even when automated systems are involved. Simply inserting technology into a process does not make the service better: its results, limitations and use must be monitored. For the professional, explaining it clearly becomes part of the job, along with the ability to recognize when human intervention is needed. A more informed client, also thanks to AI, can ask better questions and hold received proposals to account. Advice retains its value when it can respond with verifiable reasons, without relying on the advantage of knowing tools that the other person does not master.

For those who are now sixty or seventy, adapting does not mean chasing every new application. It means being able to continue managing their assets without the evolution of services forcing a silent surrender of understanding. Artificial intelligence can contribute to this outcome when it is used to clarify, learn and prepare; it can produce a new dependence when the convenience of the answer replaces all verification. The advisor’s role is also to guide the passage between these two possibilities, respecting the person’s experience, timing and abilities. The relationship becomes stronger when the client retains the ability to understand what they are doing, why they are doing it and with what consequences, even as the tools through which they do it change.