Columbia University School of Professional Studies (SPS) launched its Master of Professional Studies in Wealth Management program to prepare professionals for an industry undergoing structural change. Tracy Schwartz, the program’s principal architect and director, recently continued her conversation with Matthias Kuhlmey, an industry leader and member of the program’s faculty. Kuhlmey teaches “Special Topics in Wealth Management: Disruptive Trends,” a course exploring the forces reshaping wealth management and challenging traditional assumptions about the future of financial advice.
Building on ideas explored throughout the program, Schwartz and Kuhlmey discuss how agent-to-agent systems could reshape access to financial advice and how advisors can still create value.
Tracy: When we spoke last year, your argument was that advisors would not lose clients to AI, but to firms that use AI well. What, if anything, has changed since then?
Matthias: My argument still holds true, but it no longer goes far enough. The initial phase of AI has largely been about assisting advisors, improving processes and creating scale benefits. The next phase will be about agency. Deployed agentic systems will increasingly be able to identify client needs and ultimately coordinate with other agents to execute solutions. We are likely experiencing the onset of an agent-to-agent (A2A) era, which will further challenge and reposition human value.
With this in mind, we now clearly see that the wealth management industry is not just facing another disruptive technology cycle but a broader transformation in which digitization first removes friction, continued democratization lowers barriers of entry and, ultimately, control shifts toward the consumer. Capabilities that once were bundled inside an advisory relationship can now be produced at lower cost or are no longer scarce. The central question is therefore shifting from which technology a firm adopts to what remains valuable when technology becomes common infrastructure.
Tracy: You have described “After Wealth Management” as a working thesis you continue to develop. At its core, what do you mean by after wealth management? Are you suggesting that the traditional advisory model as we know it is coming to an end?
Matthias: No, not at all, but we have to critically question the architecture implied by the term. Traditional wealth management assumes that a client delegates, or “hands over,” wealth to a provider, often a single provider, who then organizes expertise, products and solutions around a relatively permanent relationship. The emerging model will give clients more control over the decision process, even as they delegate more of its execution to intelligent agents.
Tracy, to be clear, a client's financial ecosystem will adapt around personal aspirations and needs rather than requiring an individual to fit an institution's segmentation and service model. Advice will be less of a contractually “set” outcome and may even become episodic. My chosen terminology “After Wealth Management” describes that transition. It is not the end of advice but the emergence of a consumer-controlled, curated and intelligence-enabled system built around a category of one, rather than industry-dictated segments.
Tracy: You place particular emphasis on A2A, or agent-to-agent interactions. What would that look like in practical terms?
Matthias: Imagine, for example, that a client's personal agent recognizes an investment need, an approaching tax event or a change in family circumstances. Within the permissions and objectives established by the client, an intelligent agent could communicate with other agents, representing the specialists needed to provide a solution. Those agentic systems or networks could gather relevant facts, identify gaps, compare possible solutions and coordinate next steps. Human involvement would remain important when judgment, contextual understanding, or additional consent is required.
This is fundamentally different from placing a chatbot or even an AI model on an advisor's website. In an active A2A ecosystem, participating agents are not simply answering questions; they are helping identify issues or needs, discover specialists, assess capabilities and may even execute pre-approved decisions. It is important to note that the point of access to an advisor may no longer be through a website or referral; it may be through a client's agent.
Tracy: If you are correct in your outlook, wouldn’t this create significant challenges for advisor businesses in the current environment?
Matthias: Yes, because it could reshape the distribution of advice before it reshapes the advice formula itself. Let me explain: Successful growth in wealth management has long depended on human referrals and centers of influence. An attorney, accountant or existing client makes an introduction, and the advisor establishes the relationship. If personal and professional agents begin identifying needs and evaluating providers directly, these introductions become machine-mediated.
Consequently, a very capable advisor could be excluded before any human conversation occurs if the associated firm is not discoverable or interoperable with agentic systems. In the future, capabilities and evidence of expertise will need to be understood by machines and humans alike. In this emerging ecosystem, brand and professional reputation will still matter but will offer less protection. Overall, advice will become more transparent and comparable.
Tracy: Does greater consumer control mean clients will no longer want a primary advisor?
Matthias: Not necessarily. The progressing democratization of financial advice creates choice, but an abundance of choice creates complexity, which increases the value of curation. Many clients will still want a trusted person who understands the entire picture. What becomes harder to defend is the assumption that every meaningful financial decision must pass through one permanent relationship or one vertically integrated institution.
The more useful question will be, who or what helps you make financial decisions? For some clients, the answer may remain a lead advisor supported by intelligent systems. For others, it may be a personal agent that pairs with human counsel when needed. The industry should not assume that clients will remain in a traditional model simply because their wealth or complexity increases; it must observe where assets are moving, how advice is being consumed and where trust is actually forming.
Tracy: If machines can produce much of the technical work, what becomes of the financial advisor?
Matthias: The role moves from manager of wealth toward advocate, counselor and source of judgment. Technical competence remains essential, but it will become less differentiating when high-quality analysis, or even intelligence, is widely available. Human value will concentrate in areas where the facts do not decide the answer: competing priorities, uncertainty, difficult trade-offs, or the meaning one attaches to wealth overall.
This is the next version of the Bionic Advice model we discussed last year. The goal is still an intentional handover between machine and human, with one or the other used where it creates measurable value. As AI absorbs more of the what and the how, an advisor's contribution moves toward the why. Future advisors may need fewer process-heavy skills and stronger human qualities, including judgment, accountability and the ability to navigate competing values. In that sense, and maybe ironically, technology could make the best advisors more human, not less.
Tracy: What is another forward-looking risk or opportunity related to the rise of an agentic ecosystem?
Matthias: One area to watch is the potential convergence of agentic systems and digital assets. Agents do not hold bank accounts or enter contracts as people do; they operate through delegated permissions inside an infrastructure provided by individuals or organizations. Programmable payments, including some forms of digital assets, may help agents compensate for data, services or the work of other agents.
If so, A2A commerce could accelerate the adoption of digital assets for reasons of utility rather than investment demand alone. That development could also force a broader reassessment of traditional assets and of what constitutes value. As digital assets become more accepted, and potentially necessary to machine-mediated commerce, firms may have to reconsider how risks and opportunities are evaluated.
In line with the overall discussion, the opportunity for advisors lies in helping clients interpret new forms of value; the risk lies in applying legacy assumptions to a changing system.
Tracy: What should firms do now if their human-machine transition is still developing?
Matthias: Begin by separating topics of technology adoption from strategic relevance. Adding another tool does not answer who controls a client’s access point or why a firm deserves to remain part of the client's ecosystem. Firms need structure, including permissioned data, that can connect with outside systems and standards that both people and machines can understand. The bigger decision is to determine how human engagement creates enough value to remain visible and worth paying for.
Most importantly, however, an advisor needs to test assumptions against actual client behavior rather than extrapolating the current model. The greatest risk is continuing to optimize a legacy system while the basis of differentiation changes. The durable strategic question is simple: When access, products and intelligence are abundant, what does my firm provide that remains valuable?
About the Program
Columbia University’s Master of Professional Studies in Wealth Management program is a 16-month online program with asynchronous instruction specially designed to accommodate working professionals. It is taught by distinguished faculty with deep applied experience in their respective fields. Additionally, it is a CFP Board Registered Program designed to help students meet the educational requirement for CFP® certification.
The application deadline for the M.P.S. in Wealth Management program is May 1. Learn more about the program here.
