Swift, a global money transfer network used by banks, has suffered gaps in security standards that have resulted in at least three breaches – in Vietnam, Bangladesh and Ecuador, according to The Wall Street Journal.
The hackers were able to gain access to Swift access codes and deliver authenticated yet fraudulent fund transfer requests. The network did not become aware of the incidents in Ecuador or Vietnam until after the fact. Hackers stole $9 million from an Ecuador lender last year and $81 million from the central bank of Bangladesh in February.
A former Swift official said the network has known its user connections are its weakest link and is in need of improvement. Leonard Schrank, who served as chief executive until 2007 after 15 years in that role, said the breaches are a big wake up call.
Swift Alerts Users To Threat
Swift, in response to the thefts, urged users in a memo Friday to improve security and reminded them they are required to inform the network immediately of any suspected fraud.
A Swift board committee is examining ways to support users’ efforts to secure systems and share information about cyber threats. The network has repeatedly said that its messaging network has not been infiltrated and that the attacks are on account of problems with users’ security.
A Swift spokesperson said the management of local systems, credentials and authorization is the customers’ responsibility.
The customer notice indicates Swift is addressing the threat, Schrank said. An additional step would be to have an “anomaly detector” that would delay suspect message traffic until it can be confirmed. Another option would be to require a stronger separation between Swift systems and banks’ networks.
Network Expands To 11,000 Banks
The network began in 1973 as a cooperative among 239 banks seeking a way to communicate about cross-border payments. The network now connects 11,000 banks and other players that sent more than 25 million messages per day on average in April.
The wide range of technologies and security levels among customers in more than 200 countries makes the security challenge worse. Every user receives a unique bank identifier code containing eight characters. A smart card on the Swift terminal at the client’s end authenticates every bank’s credentials.
Technology developments have enhanced the system’s capacity and use, but they have also made it more vulnerable. Instead of walking down a hall to a Swift terminal as they did in the past, users are now automatically connected to broader bank networks.
Under a program called “4 Pillars,” Swift focused on minimizing downtime and securing its core.
The first iteration of that work addressed avoiding system breakdowns and minimizing recovery times, according to a 2006 annual report. The second iteration focused on disaster recovery and physical security at Swift sites.
The recent threats have emerged near the network’s edges. More attacks risk damaging the network’s credibility and its utility should users have to check every order.
Steve Durbin, managing director of the Information Security Forum, a not-for-profit organization for global corporations, said the hub has a responsibility to make sure the “spokes” have the proper security in place. He said the system cannot have one of the third parties seen as a weak link since it destroys confidence in the system.
The attack on Bangladesh’s central bank exposed the first weak link. The attack resulted in an $81 million loss from its account at the Federal Reserve Bank of New York.
An official for the Bangladesh Bank who did not want to be named said the bank never changed its Swift passwords when hackers breached a computer operated by a staffer in late 2015 until the thieves got the credentials to the Swift terminal and ordered the transfers in February.
A spokesman for Bangladesh Bank, Subhankar Saha, said the bank did not know any of its computers were compromised until the February heist. As many as six to eight bank employees had access to the bank’s Swift credentials, he said. He did not comment on whether the passwords provided to the employees were changed or regularly rotated.
The two parties – Swift and Bangladesh Bank – have argued over who is responsible for the compromised part of the system. The bank is responsible for the server on the premises, Swift said. The server provides the interface for bank computers to connect to the Swift network.
The bank said Swift installed the equipment, and its operations needed technical knowledge possessed by Swift technicians.
Attempts to probe payment networks include at attack in 2009 in which hackers sent fake emails by the millions to medium-size and small businesses. The emails appeared to be from a separate U.S. transfer network that a group called Nacha manages. Recipients who clicked the malware allowed the hackers to capture the users’ credential as they accessed the bank’s website.
Additional incidents could emerge. A note indicated it knew of recent fraud at a “small number” of customers, Swift said.
Bangladesh Attackers Active
FireEye Inc., a cybersecurity firm, said the Bangladesh attackers remain active.
Bryce Boland, chief technology officer at FireEye’s Asia Pacific region, said he is seeing the same threat actors target some of the company’s customers. He said he is confident other breaches are occurring.
J.P. Morgan Chase & Co. has started to take steps to secure its Swift interface by limiting access by some employees, people familiar with the bank said. The action is part of the bank’s policy of reviewing user access to certain systems after news breaks about security threats.
Lisa Sotto, who operates the Hunton & Williams LLP law firm’s privacy and security practice, said no single entity can control another entity’s systems, but tentacles are needed throughout the network.
Featured image from Shutterstock and SWIFT.