AI agents have rapidly evolved from simple task‑assistants into autonomous systems capable of executing complex workflows, making decisions, and interacting with critical infrastructure. Whether intentionally deployed or introduced through employee‑driven “shadow AI,” these agents now operate across every enterprise environment, often without adequate oversight. This shift has created a new class of cybersecurity risk: agentic threats, where AI systems themselves become the a

What Happened Bank of Baroda suffered a cybersecurity incident caused by a compromised employee email account. This was not a core‑banking breach, the bank confirmed that its core systems remained secure. The attacker gained access to the mailbox and harvested sensitive documents stored in email attachments, including KYC forms, identity documents, audit files, and loan records. This is a classic Business Email Compromise (BEC) attack. What Was Stolen Confirmed by Bank of Bar