Article by Casper Manes
In case you've been enjoying a luddite vacation for the past couple of weeks, you might have seen some of the headlines about ransomware targeting Office 365. This is a classic example of FUD gone wild, (FUD, in case you don't know, stands for Fear, Uncertainty, and Doubt) and how a single post, whether well written or not, whether accurate or not, can be picked up and suddenly becomes gospel on the Internet. Here's a little more fact-based view of what's happened.
If you were to click through some of the headlines and tweets and reports reporting on reports, you'd find it all goes back to a single post from a single third-party security vendor known as Avanan. Avanan offers products and solutions for a variety of security challenges, including DLP, antimalware, data sanitization, and more. On June 27, Avanan author Steven Toole posted an article on the Avanan website alleging that a "widespread attack on Office 365 Corporate Users with Zero-day Ransomware Virus" took place.
So what exactly happened?
Of course, it's all in the wording, and if you read further than the headline before deciding that the world was coming to an end and your grandpa was right when he warned you to stay away from the cloud, you'd see that this "massive attack" was detected against Avanan's customers who use Office 365. Given it's Avanan's own customers, you'd think that they could put out a little more accurate account of impacted users, but let's dig further. You might also find that the author makes some mistakes in his specifics about encryption algorithms, as well as his conversion of the 1.24 bitcoins to US dollars conversion. I have to give him a break on grammer since I am not the best at that either (did you see what I did there) but seriously, it's AES 256 and while Bitcoin to dollar conversions fluctuate like any other currency, around June 22, 1.24 Bitcoin was between US $674 and $700. The ransom note may have used an older conversion value when it was written which Toole perhaps was quoting. Silly malware authors...you couldn't dynamically update your ransom note?
As ransomware, Cerber is particularly nasty since if it gets in, it can result in significant data loss. Of course, that's not the authors' end goal. They want victims to pay a ransom, so what ransomware does is that it encrypts data using a strong key that cannot be brute forced or otherwise guessed. Cerber in particular targets a wide variety of file types, including
.gif, .groups, .hdd, .hpp, .log, .m2ts, .m4p, .mkv, .mpeg, .ndf, .nvram, .ogg, .ost, .pab, .pdb, .pif, .png, .qed, .qcow, .qcow2, .rvt, .st7, .stm, .vbox, .vdi, .vhd, .vhdx, .vmdk, .vmsd, .vmx, .vmxf, .3fr, .3pr, .ab4, .accde, .accdr, .accdt, .ach, .acr, .adb, .advertisements, .agdl, .ait, .apj, .asm, .awg, .back, .backup, .backupdb, .bay, .bdb, .bgt, .bik, .bpw, .cdr3, .cdr4, .cdr5, .cdr6, .cdrw, .ce1, .ce2, .cib, .craw, .crw, .csh, .csl, .db_journal, .dc2, .dcs, .ddoc, .ddrw, .der, .des, .dgc, .djvu, .dng, .drf, .dxg, .eml, .erbsql, .erf, .exf, .ffd, .fh, .fhd, .gray, .grey, .gry, .hbk, .ibd, .ibz, .iiq, .incpas, .jpe, .kc2, .kdbx, .kdc, .kpdx, .lua, .mdc, .mef, .mfw, .mmw, .mny, .mrw, .myd, .ndd, .nef, .nk2, .nop, .nrw, .ns2, .ns3, .ns4, .nwb, .nx2, .nxl, .nyf, .odb, .odf, .odg, .odm, .orf, .otg, .oth, .otp, .ots, .ott, .p12, .p7b, .p7c, .pdd, .pem, .plus_muhd, .plc, .pot, .pptx, .psafe3, .py, .qba, .qbr, .qbw, .qbx, .qby, .raf, .rat, .raw, .rdb, .rwl, .rwz, .s3db, .sd0, .sda, .sdf, .sqlite, .sqlite3, .sqlitedb, .sr2, .srf, .srw, .st5, .st8, .std, .sti, .stw, .stx, .sxd, .sxg, .sxi, .sxm, .tex, .wallet, .wb2, .wpd, .x11, .x3f, .xis, .ycbcra, .yuv, .contact, .dbx, .doc, .docx, .jnt, .jpg, .msg, .oab, .ods, .pdf, .pps, .ppsm, .ppt, .pptm, .prf, .pst, .rar, .rtf, .txt, .wab, .xls, .xlsx, .xml, .zip, .1cd, .3ds, .3g2, .3gp, .7z, .7zip, .accdb, .aoi, .asf, .asp, .aspx, .asx, .avi, .bak, .cer, .cfg, .class, .config, .css, .csv, .db, .dds, .dwg, .dxf, .flf, .flv, .html, .idx, .js, .key, .kwm, .laccdb, .ldf, .lit, .m3u, .mbx, .md, .mdf, .mid, .mlb, .mov, .mp3, .mp4, .mpg, .obj, .odt, .pages, .php, .psd, .pwm, .rm, .safe, .sav, .save, .sql, .srt, .swf, .thm, .vob, .wav, .wma, .wmv, .xlsb,3dm, .aac, .ai, .arw, .c, .cdr, .cls, .cpi, .cpp, .cs, .db3, .docm, .dot, .dotm, .dotx, .drw, .dxb, .eps, .fla, .flac, .fxg, .java, .m, .m4v, .max, .mdb, .pcd, .pct, .pl, .potm, .potx, .ppam, .ppsm, .ppsx, .pptm, .ps, .r3d, .rw2, .sldm, .sldx, .svg, .tga, .wps, .xla, .xlam, .xlm, .xlr, .xlsm, .xlt, .xltm, .xltx, .xlw, .act, .adp, .al, .bkp, .blend, .cdf, .cdx, .cgm, .cr2, .crt, .dac, .dbf, .dcr, .ddd, .design, .dtd, .fdb, .fff, .fpx, .h, .iif, .indd, .jpeg, .mos, .nd, .nsd, .nsf, .nsg, .nsh, .odc, .odp, .oil, .pas, .pat, .pef, .pfx, .ptx, .qbb, .qbm, .sas7bdat, .say, .st4, .st6, .stc, .sxc, .sxw, .tlg, .wad, .xlk, .aiff, .bin, .bmp, .cmt, .dat, .dit, .edb, .flvv
and encrypts them using AES 256. It includes a ransom note and audio that directs victims to download a TOR browser to access a particular website for directions on how to pay the ransom. It also has a time factor in it, since if users wait too long the ransom doubles. Like any other ransomware, victims really have only two choices. Pay up, or don't. If they have backups that weren't also encrypted, the choice is easy. But if they have no backups, or their backups were to a USB drive that also got encrypted, they may not be able to live without some of that data.
Cerber has ebbed and peaked a few times, and on the day in question, was detected by Microsoft, and by Google, and by Symantec, and by Avast, and by McAfee, and by lots of other vendors within a few hours. Signatures were updated, and while a handful of users on just about every messaging platform out there, whether in the cloud, hosted, or on-prem, probably opened and executed the malware and lost data, this was neither a massive attack, nor was it targeted at Office 365 users. While Microsoft has hundreds of thousands of customers with millions of mailboxes on Office 365, we have no way of estimating how many customer use Avanan's Cloud Security Platform. They do, and they didn't provide hard numbers in their post. They did estimate 57%, but heck, that could be 24 of their 42 customers or 4800 of their 8400. Only they know. Read into that what you will, but I smell something, and it's not what the Rock is cooking.
The post cites that this was detected by Avanan's Cloud Security Platform, which, as it turns out, appears to be a security and messaging hygiene platform that Avanan offers its customers. In other words, this attack that Avanan detected was not against Office 365 at all, but rather was detected by their messaging hygiene system that lives 100% outside of Office 365 and provides filtering of inbound emails to any messaging platform, including Office 365, Google, Box, and in their own words, "Any SaaS." Since Avanan can only detect what is processed by their customers, it's just as reasonable to assume that the attack targeted Avanan customers, but that's not really the case either.
What to do to stay extra secure on 0365?
So, let's consider what Office 365 customers can do. Office 365 Exchange Online Protection offers layered protections against spam, phishing, and malware attacks, and uses multiple antivirus engines to scan inbound and outbound messages for malware. This is available to every Office 365 customer that uses their EOP system for hygiene. But like any signature based anti-malware solution, including Avanan's Cloud Security Platform, it can only detect malware for which a signature exists. Until the various antimalware vendors detect and update their signatures, zero-day malware can get past this. Incidentally, Avanan's Cloud Security Platform and Exchange Online Protection both use some of the same antimalware engines to scan for malware, so don't think either is better than the other here. However, customers who want to get the maximum protection possible against zero-day infections can purchase the optional Advanced Thread Protection, which goes further than signature-based scanning by actually opening attachments in a sandbox to see what they do. If they are harmless, they are delivered. If they do anything they should not, like try to access protected memory or drop files in System32 or write to the registry or modify other data files or anything else malware does, they are blocked. While nothing is flawless, this is about as good as it can possibly get unless you want to block all attachments, period. That's a great idea, but breaks a lot of the functionality businesses use email for.
So, what can anyone do, whether they are Office 365 customers or not? How can you protect yourself from both malware, and FUD? Keep your systems patched and up to date. Use a solution like GFI LanGuard to automate this process so you have no more excuses when a patch becomes available.
Run antivirus software on all endpoints, and every intermediate system that touches data. Use different vendors to get a more layered coverage or opt for solutions such as GFI MailEssentials, for your mail security, that use different antivirus engines within one product.
If you do use Office 365, subscribe to the Advanced Threat Protection. If you use something else, push your vendor to offer similar, or consider switching. Implement SPF, DKIM, and DMARC and block anything that fails. Make sure your partners do the same.
Keep backups of critical data that are not constantly accessible to endpoint systems that could get infected by malware.
And most important of all, train your users not to open attachments from unknown senders! And whatever you do, don't read a series of amplifications of a single report from a vendor trying to sell you security solutions and assume that it's the gospel. Get validation from somewhere else, anywhere else, or remain skeptical until you do. FUD, like so much of the hate that is dominating the media lately, can get a life of its own. Don't buy into it.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.