Oi, bank manager. Only you've got my email address - where're these TROJANS coming from?
Santander scratches head over mystery malware barrage
Santander customers are continuing to complain about receiving trojans and other junk to email addresses exclusively used with the bank. The reports began last month, prompting promises of an investigation by Santander. It's still unclear whether email addresses leaked from the bank or one of its affiliates.
Independent experts said that fingering the source of this type of leak can be hard to determine.
We first heard of problems from Reg reader “Paul” in mid November. "In the last few days, I've started to receive a number of spam/virus emails sent to a unique email address only given to and used by the Santander bank to contact me," he told us.
Checks at the time revealed that Santander had already launched an investigation in response to similar reports, as SC Magazine reported in early November.
A trojan was being sent to private email addresses that, according to complaints and reports, should only have been known by institutions including Santander Bank, the UK Government Gateway and NatWest's FastPay service. The attacks were first detected by Belgian security firm MX Lab.
A substantial sample of the offending emails contain reference to a supposed "Direct Debiting Seminar Invite" and a trojan in an attached ZIP file, as explained in a blog post by MX Labs here.
Attacks against unique email addresses registered with Santander have continued into December, with another Reg reader reporting the problem a few days ago. It's unclear whether or not this is the second wave of the same attack.
Reg reader “Andrew” told us: "It appears that Santander may have had a data-breach: the customised unique email address I gave exclusively to them is now being used to send me junk email, trojans."
According to a thread on UK financial advice website Money Saving Expert, Andrew and Paul are far from the only one to run against this problem. There's been a steady stream of reports on the issue since mid-November.
There's no suggestion that any of the bank's more sensitive systems are leaking, but those who submitted posts to the MSE thread are worried about the junk mail they're receiving after email addresses supplied to the bank somehow leaked out.
In a statement, Santander told El Reg that it's continuing to investigate a possible data breach involving email addresses supplied to them:
Our investigation is ongoing. If, when it is completed, a breech [sic] has occurred we will follow the correct procedure in reporting this to the relevant authorities.
These kind of problems crop up regularly and are far from limited to Santander. It's a familiar story: someone receives a malicious email address in a unique email address box only used to register an account with one entity (typically a bank). They cry foul and tell us the bank must have had its email database breached. This is a regular theme of items in Vulture Central's mailbag.
How – and more importantly, why – could this have happened?
So what might be happening? One possible explanation is that the bank supplied its email address database to a third-party marketing affiliate and the information leaked from there. Other possibilities include that spammers are emailing address lists generated from combinations of banking-related words, so (for example) firstname.lastname@example.org happens to be hit.
Martijn Grooten, Virus Bulletin's anti-spam test director, said that leaks from third-party affiliate marketing firms are the "most likely" scenario even though other possibilities can't be ruled out. "Sending (a lot of) email isn’t trivial, what with all those spam filters out there, so almost everyone outsources it," Grooten explained.
Yet spammers hitting on the correct mailing list using some type of brute-force attack is "pretty unlikely to occur in practise", according to Grooten.
"I know non-existent addresses do receive spam, but that seems more a case of those selling lists of email addresses adding some fake ones to increase the volume, than spammers just trying everything@the_domain," Grooten told El Reg. "That’d be a huge waste of resources – and I’ve never seen it happen."
Grooten suggested other alternative possibilities that hadn't immediately occurred to us, such as recipients of spam making a mistake themselves that exposes their unique email address to unwanted attention. These mistakes can take multiple forms.
"Perhaps they misread the email address the email is sent to," Grooten explained. "Perhaps they did give it out to someone else, but have forgotten since. Perhaps the email isn’t malicious, but incorrectly flagged as such by a spam filter. Perhaps they fell for a phishing scam and filled in the email address."
"And while it’s bad if a list of addresses of a bank’s customers leaks, the fact that a Barclays customer gets spam targeting Barclays customers on an address only known to Barclays, doesn’t mean that the fact that they are a Barclays customer gets leaked: it could well be that only the address leaked and that they happen to get Barclays-related spam," he added.
Paul Wood, manager of cyber security intelligence at Symantec, agreed that it's more likely to be a third party contracted to do mailing that’s been compromised than the bank itself.
"The other possibility [that someone simply made a mistake and had actually shared the address more widely than they thought] is also a good thing to remember," Wood told El Reg. "However, that becomes less likely as more and more people report the same behaviour. This is often how many security problems become public."
Wood, like Grooten, doesn't think the torrent of malicious spam against Santander customers is based on a brute-force attack but on a list somehow obtained by crooks. How malware flingers got hold of this list remains wide open to speculation.
"If these emails are just being spammed from some botnet, then the addresses are likely to have been harvested somehow, as it's unlikely they were programmatically generated," Wood told El Reg. "But if they are sent from the genuine IP range of the supplier, then it may suggest their client's login/account has been hacked. If the attackers know enough about how the system works, all it takes is one weak password." ®