Your next financial planner might be a robot

Large Language Models like ChatGPT have been used to write everything from wedding vows to fragments of long-lost plays by Shakespeare. And soon, banks may even be using them to write your financial plan

Driving the news: Global names in financial planning, including JPMorgan Chase and Goldman Sachs, are experimenting with ways to use Large Language Models in the advice they give clients. 

  • In the United States, JPMorgan Chase may become the first major bank to release a Large Language Model AI that helps clients choose appropriate investments for their portfolios. 
  • JPMorgan Chase has filed a patent for a program called IndexGPT, which the bank claims can assist with the “selection of financial securities and financial assets” – not to mention “advertising,” “marketing,” and mundane administrative tasks. 

Why it’s happening: Robo advisors have played a role in the financial planning offerings of banks and other institutions for a number of years already, but the viral popularity of ChatGPT has reignited interest in the ways AI programs can help clients make money decisions – and companies want a piece of that action

  • Generative AI, including Large Language Models, could help banks save anywhere between $200-340 billion annually, according to one estimate.

Why it matters: As AI like ChatGPT becomes more common in every aspect of life, including financial planning, it has the potential to change the type of advice you receive – and the quality of it, too. 

  • ChatGPT and other LLMs often produce “hallucinations,” long strings of incorrect information that might appear factual or accurate if you aren’t familiar with the subject matter.
  • There are also concerns about information security and privacy, as Large Language Models harvest data to produce answers that are more and more accurate. 

Bottom Line: As with anything concerning your money, we’d recommend a “buyer beware” approach when it comes to taking financial advice. AI tools might help banks save on labour costs, but it’s TBD whether they’ll provide better advice than humans.