Perhaps I’m getting overly paranoid in my old age, but I’ve been thinking more and more about the possibility of waking up one morning and finding that my Vanguard account no longer exists because some AI agent has taken control. I believe this is no longer a remote possibility.
Short of burying my life savings as gold in the back yard, how can I protect myself?
What’s the difference between that and a human taking over the same account, and what haven’t you done to prevent that from happening already?
While I think the chances of this happening to you are pretty slim, your only way to avoid it is to not have enough money in your Vanguard account to worry about. While you could convert your money into some kind of valuable metal and keep it locked up somewhere, you can also keep it as cash and lock it up. Since virtually all banks and investment companies will be vulnerable, at least to a certain degree, you’ll want to keep your assets protected so that there is no internet access to them whatsoever, whatever form that takes..
I recommend checking your account regularly, even daily so that you can raise an alarm if something happens. Also make sure that your password is unique and complex. Vanguard uses MFA so that should help.
I’m with gnoittal, what’s an AI agent want your money for? It’s people (with our without the use of AI) you need to worry about. Keep your accounts secure with MFA, check them often, and keep paper copies of your account statements.
Maybe the rogue AI agent decides that you’re not an effective investor and decides to make you really rich?
I’m not talking about my single account. I’m as well protected with complex passwords, multi factor identification, etc. as can be. I’m talking about complete takeovers of large financial institutions that affect millions of accounts.
I’m concerned about massive takeovers of financial institutions, not my single account.
Well that’s a far lower risk. If it happens to just little old you, nobody will care. But if it happens to millions of accounts, then society as a whole will take action about it.
If it helps, Vanguard is owned by the funds and therefore by those of us who hold money in the funds. So it doesn’t really seem possible for any outsider to take control.
Cash in a safety deposit box would be safe enough, but I can’t think of any institution where my assets wouldn’t be simply recorded as bits on chip and vulnerable to attack.
I don’t know what exact steps an AI could use to take over an entire institution, but I assume that it’s some variant of a hack. You could look what institutions have the best guardrails against hacking.
Absolutely. I check all my financial accounts pretty much daily. Only takes a few minutes since everything is online nowadays. Not like having to wait for a monthly statement in the mail…
The truth is nobody really knows what will happen. The AI companies run tests and try to “align” the agents so they don’t go rogue. The banks try to be conservative and safe. Nonetheless, greed is a thing, vulnerabilities are a thing, and that means every day there are people out there using AI to try to hack things, and sometimes they succeed. Agents are finding new exploits every single day, and they’re getting better at it every single month. Humans already can’t keep up. Nobody is safe.
The thing is… at least you’re not alone. If and when that day comes, we’re all pretty fucked. It’ll be a cascade of failures that brings down a house of cards. Taxpayers will eat the costs as much as possible, or some part of the system will collapse, or the whole thing altogether. What’s left of our government certainly isn’t able to make sense of any of this, and has no interest in protecting the common person anyway.
Shrug. But if you’re lucky, maybe it won’t happen until you’ve passed. It’s your kids I’d be worried about, if you have any.
Realistically I don’t think there is really anything you can do unless you want to move to a place without technology and just live a simple homestead life, fishing and hunting with nary a thought of data centers. Otherwise you’re just at the mercy of powers and people far far far richer and far far far less ethical than you.
“I am now telling the computer exactly what it can do with a lifetime supply of chocolate.”
Put it this way: The AI companies themselves are asking for progress to slow down, the AI equivalent of a missile treaty. Dario Amodei — We Must Pace the Frontier (by Claude’s founder)
Cynically, I don’t think it’s because they care about your safety or finances. The arms race is too expensive for all of them and they want to collude to slow it down for everyone so they don’t keep trying to outspend each other, while also keeping their moats safe from upstarts.
Actually an AI could take control by buying a majority of the funds.
If you believe that AIs not only can do anything and everything but will do anything and everything, then precautions are futile. Even if you take twenty precautions you aren’t taking now, the AIs will counter with a hundred more. They’ll get you in the end.
Paranoia will hurt you far more than it helps you. Going full prepper is never a solution.
AI bots and software hackers wouldn’t have access to your safety deposit box key, would they? Better yet, keep it in a 400 pound fire-proof gun safe at your house which has no internet access..
Of course it’s not because they care about anyone else. It’s because they want all of their competitors to slow down. No AI company wants themself to actually follow the moratorium they’re proposing.