If the credit card company gives access to their account records to AI agents, that’s a vulnerability. Following tech news you see stories all the time of agents being tricked to go outside their parameters or give access to info they shouldn’t. All these companies are rolling out AI features prematurely because they are over-leveraged in AI investments that aren’t bringing returns fast enough for investors. I’m not saying the idea couldn’t have merit if rolled out properly, but we have all seen these companies rushing out half cooked products and pushing them on customers with auto opt in policies. Microsoft, Apple, Google, etc. Do you really trust Visa to do a better job at cyber security or be more cautious about integration and implementation that protects customers? I am not trying to spend hours on the phone fighting bull shit charges these sloppy AI agents “accidentally” purchased on my behalf or scammers tricked agents into making. AI Agents are getting tricked into sending password resets on accounts scammers don’t own all the time. I read the article, I understand the concept, I don’t trust these businesses to implement it responsibly. My comment was I don’t want auto opt in and I don’t see that as an uncommon outcome in the landscape.