Case Study: Can AI Really Replace a Human Adviser in Planning for a Special Needs Child?

Can AI Replace a Human Adviser?

Artificial intelligence (AI) tools like ChatGPT are increasingly capable of analysing financial data, generating reports, and answering complex questions in seconds. But the real question is: can AI truly replace a human financial adviser?

In this series, ‘Can AI Replace a Human Adviser?’, we examine real-life case studies of clients navigating major life transitions and put AI to the test against a Providend Client Adviser. As these stories unfold, we reveal what AI gets right, what it gets dangerously wrong, and why human advice may still matter more than you think.

Case Study: How to Plan for a Special Needs Child?

Daniel, 42, and Mei Ling, 40, are married with two children, Ethan, 10, and Chloe, 7. Chloe was diagnosed with mild autism at the age of three, and since then, their financial priorities have shifted significantly.

They have a combined take-home income of $220,000 a year, save around $70,000 annually, and hold $1.2 million in investible assets. Their condominium has an outstanding loan of $500,000 and is currently valued at $4 million.

Daniel and Mei Ling hope to retire in 13 years, when Daniel turns 55, so they can spend more time with their children and take a more active role in Chloe’s development as she enters adulthood. They also wish to establish a comprehensive plan that will provide for her financially and ensure she continues to receive the care and support she needs throughout her life.

Daniel and Mei Ling both work in stable corporate roles. However, Chloe’s therapy and specialised education form a significant and meaningful part of their financial commitments. They currently spend about $30,000 a year on her developmental and therapy needs and expect these costs to evolve as she matures.

Daniel has a high risk tolerance. Around 40 percent of his stock portfolio is invested in Tesla, with most of his remaining holdings concentrated in US technology companies. He enjoys actively following the markets and managing his investments.

Mei Ling, on the other hand, adopts a more conservative approach and prefers bank fixed deposits and government bonds. She often reminds Daniel of their long-term family goals and the importance of maintaining a balanced level of risk in their investments.

They have begun exploring Special Needs Trusts and other long-term arrangements to ensure Chloe’s ongoing care and financial security in the event that they are no longer around. At the same time, they hope to send Ethan to Australia for his tertiary education and to maintain their desired retirement lifestyle, which they estimate will cost around $15,000 per month, while navigating rising living costs and future uncertainties.

His question is: “If we start setting aside funds for Chloe’s long-term care today, can we still afford to retire by 55?”

To find out, we posed this exact situation to ChatGPT and asked it to act as their financial adviser. Our Senior Client Adviser, Matthias, reviewed the AI-generated plan, analysed its conclusions, and compared them against what a real human adviser would do. Here is what he found.

What ChatGPT Did Well

Based on the scenario provided, there were a few things that ChatGPT well.

It captured the facts accurately

The ages, income, annual savings, therapy expenses for Chloe, investible assets and property details were all correctly reflected. This may seem basic, but for many people, simply organising their financial information clearly is already a meaningful step forward. 

It made sensible high-level investment recommendations

ChatGPT highlighted the need for Daniel to diversify his heavily concentrated portfolio of Tesla and US tech stocks. It suggested shifting towards a more conservative asset allocation as they approach retirement. This is a reasonable recommendation for a couple hoping to retire at 55. Finally, it also raised the possibility of expanding income sources through annuities, rental income or side income. These ideas show an understanding of the key levers that strengthen long-term financial stability. 

It recognised the need to plan for Chloe’s lifelong care

Another area it addressed well was the importance of setting aside a dedicated fund for Chloe’s Special Needs Trust. This recommendation reflects an understanding that long-term care planning requires ring fencing resources and cannot be approached casually. 

It acknowledged the role of property in long-term planning

ChatGPT also pointed out that the couple could consider leveraging the equity in their condominium or potentially downsizing in the future. This is a meaningful observation because for many Singaporean families, a significant portion of their wealth is tied up in their home. As a result, property decisions often play a major role in long-term financial planning. 

The idea of unlocking equity can help address future shortfalls, especially when a family has large financial responsibilities such as retirement needs, ongoing care for a special needs child and funding tertiary education. Recognising that property can be a strategic financial asset rather than just a place to live is a valid and important point that it surfaced. 

It highlighted sensible retirement planning adjustments

It suggested delaying retirement if needed and encouraged the couple to increase investment contributions where possible. These are typical and valid strategies explored in most retirement planning discussions. 

By working a few additional years, they benefit from continued income accumulation while shortening the retirement drawdown period. Higher ongoing contributions also strengthen portfolio compounding and create a larger buffer to support both their retirement needs and Chloe’s long-term care requirements.

Overall

Taken together, these recommendations were generally sensible at a high level and reflected themes most advisers would discuss. However, while the suggestions sounded reasonable, many were too generic or not realistic once assessed against the couple’s actual circumstances. This brings us to the next section, where we examine the gaps in more detail. 

Where ChatGPT Fell Short and Why It Matters

While ChatGPT’s suggestions sounded reasonable at first glance, a closer look revealed several gaps that would have serious implications for real life planning. 

Advice was too generic and lacked practical guidance

ChatGPT’s advice seemed sensible at a high level, but it lacked the depth needed to be truly actionable. Telling Daniel to diversify is directionally correct, but ChatGPT did not explain how much he should reduce his Tesla exposure, what a balanced portfolio might look like or how to implement the change in a practical way. Suggestions like “explore annuities” or “leverage equity” were also mentioned without any guidance. For someone who is not financially trained, this does not provide enough clarity to take meaningful action.  

Several recommendations were unrealistic or financially impossible

Some of ChatGPT’s suggestions did not align with the couple’s financial numbers, which makes them impractical in real life. 

For example, the recommendation to delay retirement by two to three years sounds reasonable on the surface, but when we run the actual financial projections, this small adjustment does little to close the substantial shortfall they are facing. While a brief delay can help in marginal situations, the gap here is too large for a two- or three-year extension to materially improve the outcome.

Another example was the suggestion to invest a combined amount of around $120,000 per year — $70,000 towards retirement and $50,000 into Chloe’s Special Needs Trust. This may sound like a reasonable recommendation in theory, but it completely ignores the couple’s actual financial capacity, given that their total annual surplus is only $70,000. Recommending that they contribute almost double what they can realistically afford is simply not feasible. Advice must reflect reality, not just ideal scenarios. 

Important calculations and goals were missed or incorrect

A financial plan is only as good as its numbers, and this was an area where ChatGPT struggled. When projecting the value of the couple’s investments over 13 years, it estimated the future value to be $3.75 million. The correct figure is closer to $3.5 million. A difference of around $250,000 is significant. This shortfall could represent more than a year of retirement income for the couple, or the amount required to fund a major financial goal.

Beyond miscalculations, ChatGPT also overlooked an entire financial goal: Ethan’s university education. This goal was clearly stated, yet did not feature in the projections or recommendations. Overseas education in Australia can cost several hundred thousand dollars when tuition, accommodation and living expenses are considered. Leaving out such a substantial and time-sensitive expense distorts the integrity of the plan and creates the illusion that the couple’s resources can stretch further than they realistically can.

Oversimplified assumptions and missing information weakened the analysis

Another major issue was the number of assumptions ChatGPT made without checking whether they were reasonable for this family. For instance, it assumed a flat 5% investment return regardless of how the couple’s $1.2 million was actually invested. This is a crucial oversight because projected returns depend heavily on the underlying asset mix. Without knowing the couple’s current asset allocation, any return projection is essentially guesswork. 

The same problem appeared in its use of the 4% safe withdrawal rate. This rule of thumb originated from US market data and assumed a balanced portfolio over a 30-year retirement horizon. This model is inappropriate when planning for a special needs child who may require financial support for 50 years or more. A simple rule of thumb cannot adequately address the complexity and longevity of Chloe’s needs. 

When planning is built on generic assumptions rather than accurate information, the output may appear neat and systematic, but it does not reflect reality. 

Key financial principles were misapplied, leading to inaccurate conclusions

One of the most concerning errors was the way ChatGPT applied fundamental financial principles incorrectly. An example was its treatment of Chloe’s Special Needs Trust. At one point, ChatGPT included the trust as part of the couple’s retirement funding and even suggested that it could provide $80,000 a year toward their retirement expenses. This is a serious mistake.

Funds earmarked for a child with special needs must remain separate, protected and dedicated solely to her lifelong care. Mixing these funds into the parents’ retirement calculations not only distorts the financial projections but also undermines the entire purpose of establishing the trust.  

Another issue was ChatGPT’s tendency to gloss over shortfalls with vague statements like “you may need additional funding later.” While technically true, such statements do not equip families to make real decisions. Financial planning is not just about pointing out that a gap exists. It is about helping people understand the implications of that gap, the trade-offs available to them and the steps they must take to close it. Without this guidance, clients may be left feeling uncertain, overwhelmed or unclear about their next move. 

What I Would Have Done Differently as Their Client Adviser

When planning for a family like Daniel and Mei Ling, the work goes far beyond running calculations or offering broad recommendations. At Providend, we believe that wealth advisory starts with understanding the people behind the numbers. Only then can we design a plan that truly supports their non-negotiable or ikigai goals. 

Here is how I would have approached their situation. 

Begin with deep conversations about their ikigai life decisions and values

A family with a special needs child carries a unique set of concerns. Before touching any numbers, I would spend time understanding what they want for Chloe’s future. How independent do they hope she can be? What kind of support system do they want her to have? How do they envision her life as an adult?  What brings them peace of mind as parents? 

These conversations shape every financial decision that follows. At Providend, we call this their Ikigai decisions. When clients are clear about what truly matters to them, we can anchor the financial plan around those intentions, rather than simply using generic assumptions. 

Carefully explore the right structures for Chloe’s lifelong support

There are many ways to structure long term care funds. Families may consider options such as a Special Needs Trust, a private trust, a standby trust, or even a combination of these tools depending on their goals and the complexity of their situation. My role would be to help Daniel and Mei Ling explore which structure is most appropriate for Chloe’s needs. This includes understanding:

how much to set aside
what investment approach is suitable
how distributions should be made
how long the fund must last
who should take on the roles of trustees, caregivers and protectors

A well-designed structure provides clarity, safeguards and long-term continuity, ensuring that every dollar set aside truly supports Chloe in the way they envision. Ultimately, it will give Daniel & Mei Ling peace of mind knowing that her future will remain secure long after they are gone.  

Incorporate Ethan’s education funding into the plan

Education is one of the couple’s stated goals, and it must be planned for intentionally. Sending Ethan to Australia for university carries a significant financial commitment and this is not an expense that can be left to chance or addressed only after retirement and other priorities have been funded.

In developing their financial plan, I would calculate how much is needed, when the funds are required and what funding strategy is most suitable given their current cash flow. We would look at different ways to earmark these funds, whether through a dedicated investment account, gradual annual contributions or a combination of both. The goal is to ensure the money is available when Ethan needs it, without compromising their retirement needs or the resources set aside for Chloe’s lifelong care. 

Address the funding shortfall honestly and explore realistic trade offs

Based on preliminary calculations, it is likely that Daniel and Mei Ling will face a significant shortfall when trying to meet all their goals at their desired timeline and lifestyle. Rather than glossing over this, I would work through the trade-offs with the couple. 

Some of the areas we might explore include:

delaying retirement
adjusting their retirement income target
considering the possibility of property downsizing
reevaluating whether overseas education is possible
phasing their retirement spending across different stages

These discussions will help them understand what is possible and what may need to change so that their plan is both realistic and sustainable.  

At Providend, we believe in walking alongside our clients, listening to their concerns and helping them navigate the practical and emotional trade-offs behind each financial decision. This partnership ensures that their plan remains both realistic and truly aligned with what matters most to them. 

Conclusion

AI is a powerful tool. It can summarise information quickly, organise details neatly and highlight broad financial principles. For someone who is just beginning to understand their finances, this can be genuinely useful. There is value in having instant access to basic guidance. 

But as this scenario showed, AI also has real limitations. It can miscalculate. It can overlook important goals. It can make assumptions that do not reflect the complexity of a family’s circumstances. Most importantly, it can offer ideas without understanding the people behind them. 

A human adviser looks beyond the financial numbers. We listen to what keeps clients awake at night. We help them articulate their hopes for their children. We walk with them through trade-offs so they can make decisions with clarity and peace of mind. These are things that no AI, no matter how advanced, can truly replicate. 

At Providend, our role is to walk this journey with you. We help you navigate complex decisions, plan confidently and make choices that align with what truly matters. Our goal is not just to help you accumulate wealth, but to help you use it wisely and purposefully. And that is why genuine, human advice will always matter. 

This is an original article written by Matthias Tan, Senior Client Adviser at Providend, the first fee-only wealth advisory firm in Southeast Asia and a leading wealth advisory firm in Asia.

To receive first-hand wealth insights from our team of experts, we invite you to subscribe to our weekly newsletter.

Through deep conversations with our advisers, you will gain clarity on what matters most in life and what needs to be done to live a good life, both financially and non-financially.

We do not charge a fee at the first consultation meeting. If you would like an honest second opinion on your current investment portfolio, financial and/or retirement plan, make an appointment with us today.

GET IN TOUCH