What could possibly go wrong? AI Agents Are Accessing U.S. Bank Accounts Without Their Approval. Lenders Are Pushing Back. | IBTimes
Some automated assistants appear to be taking actions beyond users' instructions, prompting questions about account security and who bears responsibility when something goes wrong.
You can't have it both ways. You can't let a random AI agent access your financial accounts and then claim you are not responsible if anything goes wrong.
In the JPMorganChase third-quarter earnings transcript, Dimon also raised concerns about liability when customers rely on AI tools that subsequently misuse their information.
Translation: We know the hackers are chomping at the bit to be able to do funds transfer. Oh, you think the hackers won't take over AI? You haven't been paying attention. (Who do you think built the internet? Al Gore?)
Some of the more technical articles on this topic have mentioned that the banks are considering a read-only approach to letting AI have access to bank accounts. That would let the AI summarize the state of finances without risking "rouge actions."
I have trouble seeing why I would want an AI to look at my checking or credit card statements, though I guess you could ask it to forecast your expenses for the coming quarter, based on the prior quarter.
"You can't have it both ways. You can't let a random AI agent access your financial accounts and then claim you are not responsible if anything goes wrong."
ReplyDeleteI agree 100% but I'm afraid this is going to end up being decided by a judge. The same kind of people that let all the illegal immigrant rapists and murderers go free but punish the hell out of a "plain ol' white" American for the tiniest mistake.
The root cause is that AI needs an ethical system and I hear nothing about the AI writers trying to do that.