On Friday, November 6, three major Malaysian banks issued back-to-back "service disruption" notices: Bank Simpanan Nasional (BSN), Bank Islam, and Bank Rakyat. Cross-bank ATM withdrawals, mobile banking apps, and corporate websites were all hit. For customers, the day's experience boiled down to one image — three banks' ATMs, online banking, and portals all "in red" — and it fueled public worry that the country's banking infrastructure may have suffered a large-scale failure, or even a coordinated cyberattack.
The telling detail: the three failures had three different root causes. BSN attributed the disruption to high ATM transaction volumes, warning that withdrawals through the MEPS interbank network at other banks' ATMs could fail, and asked customers to temporarily use BSN's own ATMs. Bank Islam's problem was software compatibility: its mobile banking app was incompatible with Android 17, leaving customers on the new OS unable to log in or transact. Bank Rakyat's situation was the most serious — its corporate portal went down, and media reports said several banking pages, including the iRakyat online-banking login page, were defaced in a suspected cyberattack, after which the bank cut access to most of its websites.
Official accounts and public suspicion still don't line up. Bank Rakyat publicly blamed the outage on a "third-party system issue" and insisted that customer funds and data remain safe; but as of the latest coverage, none of the three banks had officially confirmed a cyberattack. More controversial was how they handled communication: all three banks closed the comment sections of their Facebook disruption notices, so customers had nowhere to ask for status updates or report their own anomalies. In the middle of a financial service blackout, several outlets flagged that move as pouring fuel on the fire.
Seen through an operations lens, the three incidents map to three classic, distinct failure classes: BSN is a capacity problem (the interbank withdrawal channel overloaded, with no anticipatory shedding or traffic steering); Bank Islam is a compatibility problem (the app release wasn't validated against a major OS upgrade); Bank Rakyat is an attack-surface and supply-chain problem (page defacement plus a third-party system failure). Three independent incidents, one shared structural gap: in every case, the failure was felt by customers before it was detected by the bank — no early warning, no automatic failover, and the only passive response was to post a notice, then close the comments.
Operations resilience for a financial institution should look different: capacity monitoring that flags interbank channel congestion hours in advance, with automatic traffic steering or graceful degradation; a release pipeline that forces compatibility validation against new OS versions; a WAF plus defacement detection that blocks malicious writes at the entry point and rolls back to a clean version in seconds. 24/7 continuous monitoring is not decoration — it is the line that holds before customers even feel pain. When three banks light up red in the same week, in three different ways, the market's assumption that "bank systems are always up" may need to be repriced.
Three banks, three different problems — capacity, compatibility, security — but one shared trait: the failure was only noticed after it had already struck. LAFA's 24/7 AI operations continuously monitors capacity headroom and version drift, and our WAF stops defacement and attack traffic at the entry point. For financial institutions, redundancy and automated failover are not a cost — they are the last line defending customer trust.