Dark Reading is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them.Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Cloud

4/10/2019
04:45 PM
Connect Directly
Twitter
LinkedIn
Google+
RSS
E-Mail
50%
50%

25% of Phishing Emails Sneak into Office 365: Report

Researchers analyzed 55.5 million emails and found one out of every 99 messages contains a phishing attack.

One in every 99 emails is a phishing attack, and a new study shows 25% of those phishing attacks bypass default security measures built into Office 365, researchers reported today.

The data comes from Avanan's Global Phish Report, which analyzed 55.5 million emails sent to Microsoft Office 365 and Google G Suite accounts. They found roughly 1% of all messages are phishing threats that use malicious attachments or links as the attack vector. Of those, 25% were marked safe by Exchange Online Protection (EOP) built into Office 365 and delivered to users.

Cloud-based email has rung in a new era of phishing, explains Yoav Nathaniel, Avanan lead security analyst and report author. "The connected nature of cloud email allows an attacker to get access to a bigger bounty from a single successful phishing attack since the credentials give them access to other connected accounts such as cloud file sharing or cloud HR," he says.

Of 55.5 million total emails analyzed, 561,947 were phishing attacks. Researchers broke the malicious messages into four categories: over half (50.7%) had malware, 40.9% were harvesting credentials, 8% were extortion emails, and 0.4% were spearphishing attempts.

Researchers scanned about 52.4 million emails directed to Office 365, of which 546,427 (1.04%) were phishing emails. They only analyzed 3.12 million emails for G Suite, of which 15,700 (0.5%) were phishing emails. In the report, researchers note how the messages were scanned after they had gone through default security but before they were delivered to users' inboxes.

They then took a closer look at how phishing emails were classified by Office 365 EOP, Microsoft's cloud-based filtering service. In EOP, emails are first sent through connection filtering, which verifies the sender's reputation and scans for malware. Most spam is deleted here, Microsoft says. Messages continue on through policy filtering, where they're evaluated against custom rules admins can create and enforce. They're also passed through content filtering, where they're checked for words and properties associated with spam. Based on settings, spam can be redirected to the Junk folder or quarantined.

After going through these layers, messages deemed benign are delivered to the inbox.

Avanan reports of the phishing emails that made it through EOP, 20.7% were marked as phishing emails and about half (49%) were marked spam. About 5% were whitelisted by admin configurations, and 25% were marked clean and successfully sent to the target user.

Bypassing Filters

How do some emails sneak through? Nathaniel says part of the reason is obfuscation, which rely on emails being displayed to end-users differently than how they appear to the machine-based security layer. Obfuscation comes in different ways: rare but legitimate email formats, which aren't properly analyzed by security but are delivered to inboxes; malformed emails and attachments that parse HTML to confuse the security layer but appear safe to the email client; and hidden characters in the email body, which are intended to trick the security filer.

Obfuscation makes up "quite a small number of attacks," says Nathaniel. "We see them targeting extremely high-profile individuals … they save it for special occasions." This may include targeting a CEO or C-suite executives of Fortune 500 companies, using attacks they don't want to land in a Junk folder.

Analysts also pulled data on different characteristics of phishing emails, which yielded some interesting data. For example, 35% of messages containing links to WordPress websites are phishing attacks. "Just the fact that it sent you a link to a WordPress site already makes the email suspicious," Nathaniel points out. And Bitcoin wallet links are almost a sure red flag: 98% of messages with cryptocurrency wallet links turn out to be malicious, researchers learned.

"It's important to note there's rarely a legitimate reason to send a cryptowallet address via email," he continues. This is typically done via text, or money is sent using an app.

Finally, the report notes out of every 25 branded emails, it's likely at least one is a phishing attempt. Microsoft is the most frequently impersonated brand (43%) for most of the year, followed by Amazon (38%), which takes the lead during the holiday shopping season.

Related Content:

 

 

 

Join Dark Reading LIVE for two cybersecurity summits at Interop 2019. Learn from the industry's most knowledgeable IT security experts. Check out the Interop agenda here.

Kelly Sheridan is the Staff Editor at Dark Reading, where she focuses on cybersecurity news and analysis. She is a business technology journalist who previously reported for InformationWeek, where she covered Microsoft, and Insurance & Technology, where she covered financial ... View Full Bio

Comment  | 
Print  | 
More Insights
Comments
Newest First  |  Oldest First  |  Threaded View
7 Tips for Infosec Pros Considering A Lateral Career Move
Kelly Sheridan, Staff Editor, Dark Reading,  1/21/2020
For Mismanaged SOCs, The Price Is Not Right
Kelly Sheridan, Staff Editor, Dark Reading,  1/22/2020
Register for Dark Reading Newsletters
White Papers
Video
Cartoon Contest
Current Issue
IT 2020: A Look Ahead
Are you ready for the critical changes that will occur in 2020? We've compiled editor insights from the best of our network (Dark Reading, Data Center Knowledge, InformationWeek, ITPro Today and Network Computing) to deliver to you a look at the trends, technologies, and threats that are emerging in the coming year. Download it today!
Flash Poll
How Enterprises are Attacking the Cybersecurity Problem
How Enterprises are Attacking the Cybersecurity Problem
Organizations have invested in a sweeping array of security technologies to address challenges associated with the growing number of cybersecurity attacks. However, the complexity involved in managing these technologies is emerging as a major problem. Read this report to find out what your peers biggest security challenges are and the technologies they are using to address them.
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2015-3154
PUBLISHED: 2020-01-27
CRLF injection vulnerability in Zend\Mail (Zend_Mail) in Zend Framework before 1.12.12, 2.x before 2.3.8, and 2.4.x before 2.4.1 allows remote attackers to inject arbitrary HTTP headers and conduct HTTP response splitting attacks via CRLF sequences in the header of an email.
CVE-2019-17190
PUBLISHED: 2020-01-27
A Local Privilege Escalation issue was discovered in Avast Secure Browser 76.0.1659.101. The vulnerability is due to an insecure ACL set by the AvastBrowserUpdate.exe (which is running as NT AUTHORITY\SYSTEM) when AvastSecureBrowser.exe checks for new updates. When the update check is triggered, the...
CVE-2014-8161
PUBLISHED: 2020-01-27
PostgreSQL before 9.0.19, 9.1.x before 9.1.15, 9.2.x before 9.2.10, 9.3.x before 9.3.6, and 9.4.x before 9.4.1 allows remote authenticated users to obtain sensitive column values by triggering constraint violation and then reading the error message.
CVE-2014-9481
PUBLISHED: 2020-01-27
The Scribunto extension for MediaWiki allows remote attackers to obtain the rollback token and possibly other sensitive information via a crafted module, related to unstripping special page HTML.
CVE-2015-0241
PUBLISHED: 2020-01-27
The to_char function in PostgreSQL before 9.0.19, 9.1.x before 9.1.15, 9.2.x before 9.2.10, 9.3.x before 9.3.6, and 9.4.x before 9.4.1 allows remote authenticated users to cause a denial of service (crash) or possibly execute arbitrary code via a (1) large number of digits when processing a numeric ...