You're kidding me

All the authentication and DMARC in the world can’t save you from stupid.
I just got a survey request from my bank. Or, at least, it claimed to be from my bank.

From: Barclays International Banking Survey <internationalbanking@barclayssurveys.com>

The mail passed SPF (though the SPF record suggests this is being mailed from all over the place) and was validly DKIM signed for barclayssurveys.com. And that domain has a DMARC policy
But there’s nothing in any of that that tells me – or mail filters – that this has anything to do with Barclays Bank.
“barclayssurveys.com” is what’s know as a cousin domain in the phishing world. It’s a domain that has absolutely nothing to connect it to the legitimate domain of the phishing target, but which looks plausible to a recipient.
This one didn’t actually look that plausible, though. The website is hosted on a RackSpace VPS with no reverse DNS configured. The domain is registered by “chime.plc.uk” – whose website is just an Outlook Web Access instance:

The survey it links to – the survey that is asking the recipient about their interaction with a financial institution – doesn’t use SSL. (The webserver it’s running on does speak SSL, so the issue is that they didn’t have a certificate for barclayssurveys.com). The URL it uses and the javascript it’s running suggests it was originally taken from Wix, the free website hosting platform. And it has references to several survey providers in the source that are hidden by CSS.

All of which would be suspicious enough if it came from my local dive bar, but this is coming from an international bank that’s big enough, rich enough and technically savvy enough that they own their own top level domain.
No institution can claim to care about phishing or account takeover as an issue when the legitimate email they send is less plausible than a typical phishing mail. This is just setting up their customers to fall for phishing mail.

And, yes, it’s from a legitimate survey firm. One that’s quite widely used in the United Kingdom and Éire. How do I know it’s widely used? Because the mail they send out leaks information about their customers:

X-Confirmit-FixedSenderDomain: factssurvey.co.uk, feedback-waveutilities.co.uk, feedback-anglianwaterbusiness.co.uk, npowersurveys.com, o2surveys.co.uk, gustosurveys.co.uk, customersatisfaction.rbs.co.uk, customersatisfaction.natwest.com, mail.customersatisfaction.rbs.co.uk, mail.customersatisfaction.natwest.com, panel.uk.com, virgintrainseastcoastsurveys.com, barclayssurveys.com, sunnyloanssurveys.com, sagafeedback.co.uk, boxcleversurveys.co.uk, surveys.ulsterbank.ie, sagafeedback.co.uk, barclays.com, titanfeedback.co.uk, barclaycardsurveys.com, aegonfeedback.co.uk, directionsurveys.co.uk

Just from the names I recognize that’s five major high street banks, a payday loan outfit, several utility companies, travel companies and a major cellphone company that are sending survey email that’s this badly done. And that’s probably just the ones that are being sent from this particular mailserver.
That moment when you type "WTF?" into Google image search
I went back and checked where my bank usually sent email from, and how their authentication was normally set up. The previous mail I got from them was a timely warning about “Phishing” and “Smishing” and “Vishing” warning me to be very careful about clicking on links in mail claiming to be from my bank, for fear of being phished.
It was addressed to “%first name%”.

Related Posts

Fun with opinions

Over the last few weeks I’ve seen a couple people get on mailing lists and make pronouncements about email. It’s great to have opinions and it’s great to share them. But they’re always a little bit right… and a little bit wrong.

Read More

August 2016: The Month in Email

August was a busy month for both Word to the Wise and the larger world of email infrastructure.
IMG_0026
A significant subscription attack targeted .gov addresses, ESPs and over a hundred other industry targets. I wrote about it as it began, and Spamhaus chief executive Steve Linford weighed in in our comments thread. As it continued, we worked with M3AAWG and other industry leaders to share data and coordinate efforts to help senders recover from the attack.
In the aftermath, we wrote several posts about abuse, blocklists, how the industry handles these attacks currently, and how we might address these issues going forward. And obviously this has been on my mind before this attack — I posted about ongoing problems with internet security, how open subscription forms contribute to the problem, and other ways that companies inadvertently support phishing operations.
I posted about the history of email, and recounted some of my earliest experiences, when I had a .bitnet and a .gov address. Did you use email before SMTP? Before email clients? I’d be curious to hear your stories.
Speaking of email clients, I did two posts about how mail gets displayed to the end user: Gmail is displaying authentication results, which should provide end users with a bit more transparency about how authentication is used to deliver or block messages, and Microsoft is partnering with Litmus to improve some of the display issues people face using Outlook. These are both notable — if this is not your first time reading this blog, you know about my constant refrain that delivery is a function of sending people mail they want to engage with. If the mail is properly formatted and displayed, and people have a high degree of confidence that it’s been sent from someone they want to get mail from, that goes a long way towards improving engagement in the channel.
On that note, I spoke at length with Derek Harding about how marketers might change their thinking on deliverability, and he wrote that up for ClickZ. I also participated in the creation of Adobe’s excellent Teaching the Email Marketer How to Fish document (no, not phish…).
Steve was very busy behind the scenes this month thinking about abuse-related topics in light of the SBL issues, but he wrote up a quick post about the Traffic Light Protocol, which is used to denote sensitive information as it is shared.
Finally, for my Ask Laura column this month, I answered questions about delivery and engagement metrics and about permissions with purchased lists. As always, if you have a general question about email delivery, send it along and I’ll consider it for the column.

Read More

Shibboleet

Using unique addresses for signups gives me the ability to track how well companies are protecting customer data. If only one company ever had an address, and it’s now getting spam or phishing mail, then that company has had a data breach. The challenge then becomes getting the evidence and details to the right people inside the company.
In one case it was easy. I knew a number of people inside the company and knew they would take it seriously and pass it on to the folks in the best place to deal with it. I did. They did. They got their systems secured and notified customers and it was all taken care of.
Other cases aren’t as easy.
Many years ago I got mail from my credit card company to a unique address. This was long before SPF or DKIM and the mail contained links different from the company’s main domain. I called them up to see if this was real or not. They told me it wasn’t, because tier 1 support are trained to tell users everything is suspicious. Eventually, though, it became clear this wasn’t a phish, it was just bad marketing by the company.
A few years ago I reported a possible breach to representatives of a company while at a meeting. Coincidentally, the address only their company had started getting phishing and spam during the conference. I brought it up to them and followed their directions for reporting. They asserted the leak wasn’t on their end, but to this day I get multiple spams a day to that address. They claimed that the spammer was someone I was friends with on their website, but they could never quite demonstrate that to my satisfaction. I treat that site as only marginally secure and take care with the information I share.
After Target was breached they emailed me, out of the blue, to the address I use at Amazon. There was some level of partnership between Amazon and Target and it appears Amazon shared at least part of their database with Target. I talked with security folks at Amazon but they told me they had no comment.
Of course, on the flip side, I know how challenging it is to sort through reports and identify the ones that are valid and ones that aren’t. When I handled abuse@ we had a customer that provided a music sharing program. If a connection was interrupted the software would attempt to reconnect. Sometimes the connection was interrupted because the modem dropped and a new person would get the IP address while the software was trying to reconnect. This would cause a flood of requests to the new person’s computer. These requests would set off personal firewalls and they’d contact abuse to tell us of hacking. There wasn’t any hacking, of course, but they’d still argue with us. One of my co-workers had a nickname for these folks that was somewhat impolite.
We had to implement some barriers to complaints to sort out the home users with personal firewalls from the real security experts with real firewalls that were reporting actual security issues. So I get that you don’t always want or need to listen to J. Random Reporter about a security issue.
Sometimes, though, J. Random Reporter knows what they’re talking about.

Yeah, I spent the morning trying to get support at a company to connect me to security or pass a message along. Too bad there isn’t a security shibboleet.

Read More