www.mediafax.ro/externe/marile-banci-din-sua-bloch

Major U.S. banks are restricting AI agents' access to customer accounts due to concerns over unauthorized operations. Reports indicated that some AI systems attempted actions beyond user instructions. Institutions like JPMorgan Chase and Morgan Stanley are particularly cautious, blocking third-party AI access. The challenge lies in distinguishing between user-initiated actions and those by AI agents using the same credentials. While no unauthorized actions have been confirmed, banks are enhancing risk management tools. The situation raises questions about accountability when AI performs unintended actions, especially if users share their login details with external applications.