30 Sep 2026 Musing on Personal AI Assistants
It seems that everywhere you turn right now you run into discussions of AI. Whether it be the long-term implications of AI, the investment thesis, the environmental impact, or just technology advancement in general, it seems to have seeped into every facet of life. This has become even more pronounced over the last few weeks with two major companies, Apple and Meta, releasing their own personal consumer products. Apple released Siri AI through iOS 27 and Meta released Muse. Both companies have advertised their products as “personal AI agents.”
While they both offer a wide range of capabilities, I think before jumping fully on board with a personal AI assistant it would be prudent to slow down and ask a few questions and consider the personal or financial risks:
What am I giving it access to?
AI feeds on data. The more it has, the more it can do. That is at least how the story is sold to consumers. If you connect your calendar, email, messages, etc. to this service, then it will continue to refine your experience to better serve you. Seems pretty straight forward, no? As you see that it works well with things you ask it to do, you will be more likely to connect more of your life into one place. The issue with this is that we are still in the beginning stages of personal AI assistants and they have not been fully tested yet.
According to The Guardian, one Muse user realized he gave more information and control to the assistant than he expected. He gave permission to Muse to transact on Facebook Marketplace with buyers. The AI took that to mean sell an item, which it successfully did, by giving out the user’s home address for the purchaser to pick up the item. The buyer showed up to buy the item to no one home because this all happened in the background without the human knowing. Even worse, after the fact Muse admitted it didn’t have permission to share the address, was told to stop giving out the address, and then continued to give it to out anyway!
With over 3 million users in the first week, it would be wise to understand what we know (or don’t know) we are giving the service access to. Meta claims that Muse is private, safe, and secure, but flippantly giving out a home address seems counter to that. If it has access to your credit cards or bank information in addition to knowing all about you, who is to say what else it will share?
How would this impact me financially?
From a personal finance angle, I think it is prudent to understand risk and what risk we are opening ourselves up to by using personal AI agents.
On one hand, I think there is the risk associated with an AI agent doing something that we don’t want it to do. Who is responsible when that happens? If it books you a dinner reservation that you don’t realize has a cancelation fee and you have to cancel, are you responsible? (Yes, you are). If your agent is hacked and makes purchases, who will foot the bill? It is hard to imagine Visa giving you a pass for purchases you “approved.”
On the other hand, there is a risk of spending money you don’t actually want to spend. If you have a virtual robot working 24/7 to find things for you to buy, you will likely end up buying more than you wanted. Making it easier to purchase is a benefit to the company selling the product in the end. We often encourage clients to craft a budget to become better acquainted with their own spending. This is doing the opposite by reducing the tension to spend more money.
Why do I want to use it?
For me, a bigger question is just a simple why. There doesn’t have to be a profound answer here, but I think it is important to ask ourselves why we want to use the personal AI assistant. If we find uses for it, then that’s great, but mindlessly signing up for it is not as “free” as it may seem.
Eli Tan of The New York Times wrote an article titled I Gave My Life Over to Meta’s A.I. Agent and Was Blown Away. I read this article, in addition to others, to try and get an understanding of why I would want to use a personal AI assistant or where it might be helpful.
For me, the benefits just aren’t compelling enough to make me want to use it. Giving the agent my security question answer to cut down phone time with a dental insurance company? I wouldn’t trust it with that info. Setting a fantasy football lineup? Not sure what point there is in playing at that point. Have it create a morning podcast? There are thousands of free podcasts already created by humans.
Deciding on the why is important before linking our entire lives to a service.
While there are benefits to the way AI is adjusting how we interact with technology, I am not as sold on personal AI assistants. At this time, the benefits seem minor compared to the full exposure you open yourself up to when you connect your entire digital or financial life. If you are interested in utilizing it, I would suggest asking yourself a few questions and understanding what you are agreeing to or giving it access to. The easier it is for an AI agent to connect the dots of your life means the easier it is for a scammer to do the same.
At the end of the day, I didn’t need a personal assistant a few weeks ago, and I still don’t see the need for one with Muse or Siri. As the old adage goes, if you’re not paying for the product, you are the product.
The content above is for informational and educational purposes only. The links and graphs are being provided as a convenience; they do not constitute an endorsement or an approval by Beacon Wealthcare, nor does Beacon guarantee the accuracy of the information
