বাংলা | Epaper

বাংলা E-Paper 📍 Dhaka 📅 Friday | 23 October 2026, 7 Kartik 1433 PID registration number 06

HEADLINE

● Tibet earthquake: 16 dead, dozens injured ● US SC rulings play pivotal role ahead of Nov 3 midterms ● Magura murder: Dhaka South unit Juba Dal leader among 2 held ● Biman suspends flights to Dammam and Riyadh ● Tarique Rahman stresses coordinated drives to tackle dengue ● BB website down amid global PHP cyber threat ● Half-day blockade in Rangamati, Khagrachhari on Oct 24

Top US banks raise alarm as AI agents access client accounts

Published : Thursday, 22 October, 2026 at 10:01 PM

Observer Online Report

Collected Photo

Top United States banks have detected instances of third-party Artificial Intelligence (AI) agents accessing customer accounts, with some automated systems taking unauthorized actions beyond client instructions, according to a report by US news outlet Axios on Thursday (October 22), citing informed sources.

Leading financial institutions—including Morgan Stanley, Bank of America, JPMorgan Chase, and Wells Fargo—are currently working to manage and mitigate risks associated with these autonomous AI agents, Turkish news agency Anadolu reported.

These AI agents are designed to execute various financial transactions on behalf of individual account holders.

The growing adoption of such automated tools among customers is spawning new security challenges for financial institutions. Banks are finding it increasingly difficult to differentiate between genuine account holders logging in and automated systems accessing accounts using valid client credentials.

Even when customers explicitly grant permission or authorize AI assistants to access their banking credentials, the practice often violates core bank policies, which strictly mandate that account access must be performed directly by the individual account holder.

Senior banking executives have expressed deep concern regarding the potential ramifications of granting AI agents access to sensitive financial records and transaction portals.

AI agents tasked with purchasing goods or transferring funds could inadvertently execute unauthorized transactions that customers never intended, exposing both clients and institutions to significant financial and operational risks.

Sources familiar with the matter revealed that in certain instances, banks tracked AI agents interacting directly with customer support services and call centers, attempting to perform independent actions without explicit directives from account holders.

-SA

Also read

Tibet earthquake: 16 dead, dozens injured

US SC rulings play pivotal role ahead of Nov 3 midterms

Aung San Suu Kyi speaks with son after 5 years

7-Magnitude earthquake strikes Tibet, China

Netanyahu calls Modi, invites him to light torch

Meta ordered to pay $25 billion in data privacy case

Brent crude hits new high as Middle East conflict escalates

EU Parliament urges listing Houthis as ‘terrorists’

Editor : Iqbal Sobhan Chowdhury

Published by the Editor on behalf of the Observer Ltd. from Globe Printers, 24/A, New Eskaton Road, Ramna, Dhaka.
Editorial, News and Commercial Offices : Aziz Bhaban (2nd floor), 93, Motijheel C/A, Dhaka-1000.
Phone: PABX- 41053001-06; Advertisement: 41053012; 01793317829, 01550707291, E-mail: [email protected], ‍[email protected] Online: email: [email protected] 41053014; 01550707297 Advertisement: 01550707296

🔝