Cyberattacks that aim to fraudulently use banks' connections tothe Swift messaging system have continued over the past few months,showing that the threat hasn't subsided after hackers used theglobal service in February to steal $81 million from Bangladesh'scentral bank.

Many of the attacks have been thwarted, Swift said in a letterto customers dated Aug. 30, either by correspondents stoppingsuspicious messages or as a result of tightened customer securityprocesses implemented with Swift's help. But others havesucceeded.

The affected banks “shared one thing in common; they have allhad particular weaknesses in their local security,” Swift said inthe letter. “These weaknesses have been identified and exploited bythe attackers, enabling them to compromise the customers' localenvironments and input the fraudulent messages.”

Natasha de Teran, a spokeswoman for Swift, declined to commenton whether any money was stolen in the latest attacks. The newattacks were reported earlier by Reuters.

Swift repeated in the letter that its core messaging service issecure and that the security of computers linked to the system isthe responsibility of member banks. It also laid out specificpriorities for how banks can protect their systems.

The group noted that customers must update their access softwareby mid-November and warned that if they failed to do so, it couldinform regulators or bank counterparties.

Bloomberg News

Copyright 2018 Bloomberg. All rightsreserved. This material may not be published, broadcast, rewritten,or redistributed.