Why banks are banning AI agents

Most of Canada's big banks are blocking AI agents from online bank accounts. RBC, TD, BMO, and CIBC say they're concerned about safety and privacy as AI agents grow in popularity. This applies to popular apps like Meta's Muse or OpenAI's Dots. More people are using these agents to shop or just get through their to-do list. They can connect to bank accounts and credit cards. Though buying things typically requires human approval. But the banks say they don't trust these digital helpers. RBC says it will block third-party personal AI agents from direct access to authenticated digital properties. That includes online banking. TD, BMO, and CIBC are making similar moves. Meanwhile, Scotiabank says it will limit AI agents from accessing its online platforms. One technology expert says the banks are right to be suspicious. I think the banks are making a smart choice. We've seen this summer a whole bunch of examples of AI agents going beyond. what they were asked to do, causing harms, hacking things, doing all kinds of different stuff. And BMO's CEO is telling customers not to give these agents account numbers, passwords, or PINs. I personally wouldn't. There's no governmental legislation that defines things like liability. If an agent does something that it's not supposed to do, we know that AI is prone to making mistakes or misinterpreting prompts from users. But this doesn't mean the banks aren't excited about AI. They've invested billions, and they're using the technology inside the bank to speed up work, like approving mortgages.