Banks stopped three new attempts to abuse the Swift financial transfer network this summer, its CEO Gottfried Leibbrandt said Monday, as he announced Swift's plan to impose tighter security controls on its customers.
Swift provides the network that banks use to exchange funds internationally, and hit the headlines in February when attackers almost got away with a billion-dollar heist at Bangladesh Bank. In the end, they only succeeded in stealing US$81 million after hacking bank systems connected to the Swift network.
That prompted Swift to ratchet up security around its systems, which weren't themselves breached, updating the software it provides banks and adding new audit and verification tools.
Despite the risk to their businesses and reputations, though, banks have been slow to make the changes, so from next year, Swift is going to make many of them mandatory.
Leibbrandt revealed details of the three new thwarted attacks during a keynote speech at the Sibos banking conference in Geneva on Monday.
A few months ago, he said, he received a call. "One of our banks had been alerted by their clearing correspondent that there was something fishy with their transactions," he said. An investigation showed that the bank had been compromised and that payment reports had been altered -- similar to the way in which the Bangladesh Bank theft was performed.
"Next day I get another call. The clearing correspondent had found that the ultimate beneficiary of these transactions, a mule account, featured in transactions of yet another bank. We contacted that bank and that bank too had been compromised," he continued.
"Fortunately, it had already been alerted by the bank of the ultimate beneficiary, who had had anti-money-laundering concerns over that final account," he said. By working together the four banks, on four different continents, were able to prevent any money from being lost.
"A few weeks later, another case," Leibbrandt said. "This was a bank that had the latest antivirus and had updated the latest security patch on our software, and in that case alerts from the antivirus and from the latest security patch triggered alerts that prevented further fraud happening there as well."
That third failed attack showed the importance of banks securing their own environments through multifactor authentication, securing their credentials, using antivirus and firewalls, and applying the latest updates and security patches to operating software, all things Leibbrandt likened to "basic hygiene."
But, he warned, "There are other cases we've seen where the attacks are successful, and we expect this to continue and get more sophisticated."
To combat such attacks, by next July Swift will require customers to self-certify their compliance with a set of 16 core security standards, and from Jan. 1, 2018 will also conduct audits.
Swift will make information about banks' compliance with the new rules available to their regulators and to their correspondent banks, allowing the latter to check whether their counterparty "has washed his hands for dinner, so to speak," Leibbrandt said, continuing his hygiene analogy.
Details of the new security standards aren't finalized yet: Swift plans to hold a two-month consultation on them from the end of October, publishing the final rules next March.
It's not just Swift's customers that will have to change to keep out the attackers.
"We are also investing in our own security," Leibbrandt said. "We have not been breached, as far as we are aware -- and I always add the 'as far as we are aware' because I truly believe that in cyber only the paranoid survive."