Wealth managers are no longer only competing with rival firms, robo-advisors or low-fee index funds. They’re now facing a stranger challenge: their own clients’ AI chatbots.
CNBC’s Inside Wealth reports that large language models are giving financial advisors a new kind of pressure, as clients use tools like ChatGPT and Claude to ask investment, tax and portfolio questions before speaking to a human.
That does not mean AI is replacing wealth managers overnight.
It means clients now have a free, fast second opinion before they enter the room.
AI is becoming the first financial conversation
For years, the human advisor controlled the first serious money conversation. They framed the client’s goals, explained the options, and shaped the recommendation.
AI changes that order.
A client can now ask a chatbot whether two ETFs overlap, whether a tax strategy makes sense, or whether a retirement withdrawal plan has obvious gaps. The answer may be incomplete or wrong, but it gives the client a starting point.
That changes the power dynamic.
So the threat is not perfect AI advice. It’s confident AI friction.
A chatbot does not need to be better than a human advisor to make the advisor work harder. It only needs to give the client enough language to challenge the plan.
We’ve already seen this shift in consumer finance through OpenAI’s ChatGPT personal finance push, where AI tools move closer to budgeting, account analysis and money management.
Why this hits the wealth model
Wealth management fees have always depended on trust. Clients pay because they believe the advisor understands risk, tax, estate planning, markets and behaviour better than they do.
But when a client can generate a clean-looking second opinion in seconds, that trust becomes more conditional.
This is why advisor platforms are adding AI into their own workflows. In February 2026, Altruist launched AI-powered tax planning inside Hazel, a platform for wealth professionals. The company said the tool can read tax documents, account statements, meeting notes and CRM data to create personalised tax strategies in minutes.
That matters because tax and planning work are among the areas advisors use to justify higher fees.
We think that is the uncomfortable part. AI may not remove the advisor. It may remove some of the mystery around the advisor’s value.
The privacy problem is just as big as the accuracy problem
There’s another issue hiding underneath the excitement: data.
Financial questions are rarely abstract. They involve tax documents, account balances, trust deeds, estate plans, debt, salaries, spending patterns and family obligations.
OpenAI’s own page on personal finance in ChatGPT frames the tool as a way to understand spending, accounts and financial patterns. But OpenAI also makes clear that ChatGPT is not a replacement for professional advice.
That caveat matters.
A person who pastes sensitive documents into a consumer chatbot may not have the same protection they would get inside an enterprise wealth platform. Altruist, for example, says Hazel uses zero data retention agreements with model providers and does not use customer data to train AI models.
South Africa should watch the advice boundary
For South African readers, the key question is simple: when does AI move from education into advice?
The Financial Sector Conduct Authority supervises financial services conduct in South Africa and gives consumers ways to verify authorised financial services providers. That matters because money advice is not just another chatbot use case.
A chatbot can explain what diversification means. But telling someone how to invest retirement savings, restructure a portfolio or handle tax exposure is a different category of risk.
This is why global AI platforms keep drawing boundaries around high-stakes use. OpenAI’s usage policies restrict certain high-impact decision-making and warn against using AI in ways that replace licensed professional judgment.
The lesson for South Africa is not that AI finance tools should be blocked. It’s that advice, accountability and data protection need to stay visible.
This connects with a wider local issue: South Africa’s fintech sector is already moving toward AI-assisted money products, digital banking and automated financial tools. That’s why South Africa’s AI-fintech moment is not just a startup story. It’s also a consumer-protection story.
Human advisors still have one strong advantage
AI can summarise. It can compare products. It can draft questions. It can even run scenarios when connected to the right data.
Anthropic says Claude is already being deployed across financial services, including research, analysis, underwriting and other knowledge-heavy workflows.
But the best human advisors still do something harder to automate: they understand the person.
That matters most during messy moments. A market crash. A divorce. A death in the family. A liquidity crunch. A business sale. A tax shock. A retirement decision that looks mathematically correct but emotionally wrong.
AI can produce an answer. A good advisor can ask whether that answer fits the client’s real life.
Still, advisors should not get too comfortable. AI tools are getting closer to the money itself, not just the conversation around money. We’ve already seen ChatGPT and Claude connect to live trading apps, which shows how quickly the gap between advice and action can shrink.
The same theme appears in agentic trading, where platforms are testing whether AI can move from analysis into execution. That’s why Robinhood’s AI trading ambitions matter beyond active traders.
FAQs
Can ChatGPT replace a wealth manager?
Not safely. ChatGPT can help explain financial concepts, compare options and prepare questions, but OpenAI says it is not a replacement for professional financial advice. Complex financial decisions still need licensed, accountable human judgment.
Why are wealthy clients using AI chatbots for money questions?
They want fast second opinions. CNBC says large language models are putting new pressure on wealth managers because clients can now arrive with AI-generated questions. That makes meetings more informed, but also more challenging for advisors.
What should South African investors watch?
South Africans should watch the line between financial education and regulated advice. The FSCA supervises financial services conduct, so AI tools used by banks, insurers or advisors will need clear accountability. The safest use of AI is to learn, prepare questions and verify information — not blindly follow investment instructions.
