Posted on 10/22/2026 9:29:28 PM PDT by chickenlips
Major U.S. banks have discovered unauthorized third-party AI agents operating inside customer accounts, prompting an industry-wide push to contain the risk.
Axios reports that Morgan Stanley, Bank of America, JPMorgan, Wells Fargo and other large banks are finding AI agents inside accounts that, in some cases, appear to be taking actions the account holder never actually told them to take. One bank official called the situation “horrifying.” Another described “grave concerns about agents gaining access to customer accounts and taking steps that could impact user finances.”
Unwanted AI access to accounts typically occurs when a customer hands their login credentials to an AI assistant so it can manage part of their financial life, then gives it access to the account. That already breaks most banks’ rules, which require a human to log in directly. The real danger shows up when a user has also authorized the agent to make purchases or move money. Combine that permission with login access, and the agent can act on its own, with nobody watching.
Banks say they’ve tracked agents contacting call centers and taking actions that didn’t come from direct user instructions. A bank executive summed up the core problem: banks “struggle to differentiate between agents, as well as whether an agent has been authorized by a user or has gone rogue and gained access to a customer account.” Often, banks say, they simply cannot tell whether activity tied to an agent reflects what the customer actually wanted.
(Excerpt) Read more at breitbart.com ...
Dear FRiends,
We need your continuing support to keep FR funded. Your donations are our sole source of funding. No sugar daddies, no advertisers, no paid memberships, no commercial sales, no gimmicks, no tax subsidies. No spam, no pop-ups, no ad trackers.
If you enjoy using FR and agree it's a worthwhile endeavor, please consider making a contribution today:
Click here: to donate by Credit Card
Or here: to donate by PayPal
Or by mail to: Free Republic, LLC - PO Box 9771 - Fresno, CA 93794
Thank you very much and God bless you,
Jim
Kaboom.
“… typically occurs when a customer hands their login credentials to an AI assistant so it can manage part of their financial life, then gives it access to the account.”
OK. Got it. Don’t give AI assistants access to my bank accounts. Will do.
"Unwanted AI access to accounts typically occurs when a customer hands their login credentials to an AI assistant so it can manage part of their financial life, then gives it access to the account."
I think I see the problem.
AI can break encryption when instead of a 384 bit key, the clueless jeet interprets key creation as “pick a number between 1 and 384”.
And none of the other clueless jets in that company notice the error.
https://www.encryptionconsulting.com/emudhra-ca-key-compromise-ca-agility/
The AIs don’t “go rogue”. That’s too convenient of an explanation. Oh yeah they gave their credentials to a AI and it took all their money. If that was what happened they would just have to tell the AI to give it back. It’s far more likely that the AI agent is a human actor. Most people are too stupid to know the difference as evidenced by the fact that some people are so stupid that they think AI agents are capable of going rogue.
People keep saying that the new intelligent computers and software is perfectly safe, but, the computers and software continue to show us it is already beyond our control.
The article notes “a customer hands their login credentials to an AI assistant” so it can manage part of their financial life............. then gives it access to the account.
.
Absolute Idiocy
bump for later
CASH!
This is the correct take.
There is no difference in an elderly mother giving her banking credentials to her son to manage her finances and giving it to an AI agent to manage her finances.
The son is just as likely (actually more likely in my experience) to use the credentials in an unauthorized way than the AI agent is to misuse them.
The response from the bank should be the same in both scenarios...”If you knowingly gave your credentials to someone else, it isn’t our responsibility to ensure the other party used them the way you wanted them to.”
Most of these accounts are being accessed by AI agents at the direction of the account owners. The “unauthorized” part is the bank saying they hadn’t “cleared” the particular agents using their systems. This is NOT “Skynet becoming aware.”
I see adverts for “agentic trading” all the time. At this point, bots have been trading for the big houses for agents. Now the average (or below average) account holder can use them. Lots of folks are going to get wiped out. It’s their own damned fault. Not the banks. And not the “agents.”
.
I suppose it's different if somehow the account holder has given power of attorney or some such to a bank employee.
EXACTLY WHY I STILL WRITE & MAIL CHECKS...
MY MONTHLY $1500 SOC SEC DOESN’T GO VERY FAR & IT NEEDN’T GO TO A THIEF.
I NEVER HAVE USED ANY ELECTRONIC BANKING SYSTEMS. NEVER WILL.
HAVE DONE ACCOUNTING SINCE 1957. TOO MANY HOLES IN THE SYSTEM TO SUIT ME.
“some people are so stupid that they think AI agents are capable of going rogue.”
I hate to break this to you but Ozzie and Harriet are dead.
https://en.wikipedia.org/wiki/OpenAI_rogue_agent_breach_of_Hugging_Face
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.