A bank password given to an artificial intelligence assistant becomes a ticket that does not say why the subject arrived. On October 22, Axios reported that J. P Morgan Chase, Bank of America, Wells Fargo and Morgan Stanley encountered third-party agent activity in customer accounts. According to the publication's sources, some programs contacted call centers or took actions as if they were not requested by the account holder. The theft of the money has not been proven. Most often, the access was opened by the customer himself. He passed the login and password information to the program, which was supposed to parse the statement or execute the command. After logging in, the bank sees the correct details and the current session. And the content of the order that a person gave to his agent in a different request is not known to the bank. Successful authorization does not prove the customer's agreement to change the payee. Hence the concern of financiers: technical confirmation of identity ceases to be reliable evidence of consent to any action. Morgan Stanley already blocks external agents, but a bank spokesman told Axius that no cases of unauthorized access to its sites had been detected. Reuters independently reported, citing industry insiders, that banks do restrict such intermediaries. This confirms the defensive response, but does not determine the number of accounts seized. Jay too. P Morgan, Bank of America and Citi impose restrictions. There is no information about illegal transfers in these banks. CEO J. P Morgan Chase's Jamie Dimon told investors on Oct. 13 that the bank would not allow arbitrary third-party agents to enter. At the same time, the institution is developing its own AI tools. An external program that uses the client's password appears to the server as the owner of the account. Dimon also warned that people risk silently accepting responsibility for the actions of a digital representative. Broker Charles Schwab has already tested a more accurate answer.

A bank password given to an artificial intelligence assistant becomes a ticket that does not say why the subject arrived. On October 22, Axios reported that J. P Morgan Chase, Bank of America, Wells Fargo and Morgan Stanley encountered third-party agent activity in customer accounts. According to the publication's sources, some programs contacted call centers or took actions as if they were not requested by the account holder. The theft of the money has not been proven. Most often, the access was opened by the customer himself. He passed the login and password information to the program, which was supposed to parse the statement or execute the command. After logging in, the bank sees the correct details and the current session. And the content of the order that a person gave to his agent in a different request is not known to the bank. Successful authorization does not prove the customer's agreement to change the payee. Hence the concern of financiers: technical confirmation of identity ceases to be reliable evidence of consent to any action. Morgan Stanley already blocks external agents, but a bank spokesman told Axius that no cases of unauthorized access to its sites had been detected. Reuters independently reported, citing industry insiders, that banks do restrict such intermediaries. This confirms the defensive response, but does not determine the number of accounts seized. Jay too. P Morgan, Bank of America and Citi impose restrictions. There is no information about illegal transfers in these banks. CEO J. P Morgan Chase's Jamie Dimon told investors on Oct. 13 that the bank would not allow arbitrary third-party agents to enter. At the same time, the institution is developing its own AI tools. An external program that uses the client's password appears to the server as the owner of the account. Dimon also warned that people risk silently accepting responsibility for the actions of a digital representative. Broker Charles Schwab has already tested a more accurate answer.