Automating financial advisor admin work

Anthropic is now taking Claude directly into the financial-advisory workflow.

Claude for Financial Advisors will be able to pull information from CRMs, custodians, portfolio reports and meeting notes; help with review preparation, follow-ups and compliance checks, while keeping advisors responsible for approving the output. Importantly, the advisor remains responsible for reviewing and approving the output.

At first glance, this looks like another story about productivity.

I think the more interesting question is what happens to the 'value of advice' when much of the information-processing around it becomes dramatically easier.

For a long time, parts of wealth management have been built around an information advantage: access to research, the ability to compare products, construct reports, synthesise market developments and present recommendations. AI will steadily compress that advantage.

But the hardest decisions we encounter with families are so because the information is incomplete, the future is unknown, and the choices are deeply personal.

Families with significant wealth never paid wealth advisors for simply “better investment ideas". In fact, the most consequential financial decisions I have seen families make rarely suffer from a lack of data.

They look like:
- Should I finally reduce my holding in the company that created most of my wealth?
- How much of our family's capital should remain liquid?
- Should we back our child's entrepreneurial venture? And if so, as parents or as investors?
- How much investment risk is appropriate when we no longer need to take much risk at all?
- Should the next generation begin participating in investment decisions now, or only after they inherit?
- When an exciting private investment comes through a trusted friend, how do we separate the opportunity from the relationship?
- During a market correction, is the portfolio actually wrong or are we simply uncomfortable?

These and many such problems require context, judgement, trust and, occasionally, the willingness to tell someone something they would rather not hear.

AI and Advisors are not on opposite sides of this transition. Used well, AI could remove a great deal of the mechanical work that surrounds advice and give advisors more time for the part that actually matters: 'thinking alongside the client.'

Advice can get more efficient. But the standard for a wealth advisor is higher: Can it help us make advice more thoughtful?

I've seen a lot of "doomer" posts lately about AI replacing advisors, or creating financial plans in minutes.

I'm not worried at all, but here's my two cents on it.

(Everything below assumes you're using a compliant and data-secure AI system)

1. Financial Planning isn't a noun, it's a verb. The posts I've seen about AI doing financial plans in seconds-flat are missing the point. Trust me, as a client you DO NOT WANT a financial plan created in 10 minutes, and as a working advisor you DO NOT WANT to be able to create 10 plans for 10 clients in a day. I don't believe my financial plans really get useful until about the 12 - 18 month mark: the point where the archetype of what I think this person is *like* (executive, business owner, retiree) becomes more an understanding of who they really *are* (burning out in a job they hate and at risk of early heart attack, running a business that is entirely dependent on them and will never sell, desperately under spending because they're worried their kids won't get ahead in life).

2. Don't refuse to learn to use AI. It will save you and your team a ton of time, likely all on the tiny administrative tasks you already hate: notetaking, finding details in huge documents, formatting reports, etc.

3. Don't stop at the chatbot and notetaker - embrace the "vibe code". In college I took an intro to C++ class that essentially taught me to print "hello world" on a screen - and that's all I know about programming. Last weekend on a train I was able to code an app, using Claude, that will take a networth .CSV file from Conquest and search holdings on the internet for sector and geography data, and then spit out a pretty and interactive dashboard to show the client's net worth broken down across accounts, asset classes, and sectors. I didn't code it, Claude did, but as the planner I guided it to make the app useful... and that was just for fun on a train.

4. Clients are going to be using AI like a search engine to find you... or not... are you visible? I've now heard marketers talk about "AIEO" like "SEO" (Search Engine Optimization) so might want to read up on that. Clients will also be second-opinioning recommendations with AI (should they? Perhaps not, as the results are often wrong, but they are doing it anyway). We can't run, we can't hide, it's here. This is an opportunity to have that conversation as to *why* AI can't be relied on for wholesale financial advice.

5. Ultimately, to spin the doomer narrative around, this could be the beginning of a Golden Age of financial advice - one where administrative and data-entry burden of our time is lessened or replaced almost entirely. What fills that void? Human connection and conversation. If you were previously spending 1 hour per day transcribing notes, 1 hour per day crunching calculations and working *in* planning software, and 2 hours per day responding to emails... what would you do with 4 hours back?