Mitigating the risk of zero-day attacks against Microsoft 365

John Mc Loughlin, CEO, J2 Software.

---

Microsoft 365 servers contain extremely sensitive information and most organisations simply cannot do without them for an extended period of time. Zero-day attacks, those attacks that take advantage of a vulnerability in software or firmware found by cyber criminals before a vendor can issue a fix, are one of the cyber criminal’s greatest advantages. 

Often providing privileged access, zero-day attacks become the means for a single cyber criminal group to potentially gain access to literally tens or hundreds of thousands of organisations in every geography, industry vertical, size and country. There are even posts on the dark web by cyber criminal gangs offering as much as $3 million for a zero-day remote code execution exploit.

According to a recent analysis, there were 80 zero-day vulnerabilities exploited in 2021, a 166% increase from 2020, with Microsoft, Apple and Google products most frequently targeted.

There was a zero-day attack on Microsoft Exchange in early 2021 that originated from a Chinese cyber espionage group, dubbed “Hafnium” and, in total, exploited four vulnerabilities to exfiltrate mailbox contents, gain admin control over compromised servers and install malicious software.

While Microsoft provides a world-class layer of security around its services, as well as copious amounts of QA to ensure its products are secure from the start, the Hafnium attack is a reminder that no vendor – not even Microsoft – is impervious to very intelligent hackers who spend all their time trying to find a proverbial Achilles’ heel in a vendor’s code.

And should a zero-day attack occur, initial mitigations and updates generally are not immediately available, putting organisations that are attack targets largely helpless to defend themselves during the time between an attack’s discovery and mitigation.

But what happens if the target is Microsoft 365, specifically Exchange Online, as e-mail continues to be the number one initial attack vector? What can you do both proactively and reactively to mitigate the risk of a zero-day attack against it?

According to Mimecast, there are five key risk mitigation steps organisations can take to protect their operational investment in Microsoft 365. Considering that more than 90% of all cyber attacks start with an e-mail, it is crucial to look at this aspect in more detail.

While we’re talking about protecting against zero-day attacks, no rule says a threat actor can’t also attempt to take advantage of e-mail to spread malware within an organisation. And since this is one of those steps you can more easily address, it makes sense that you need to have a layer of protection that resides logically where an e-mail enters your organisation.

What’s needed is a defence-in-depth approach using third-party solutions that augments any built-in security on Microsoft’s part. Microsoft 365 E3 licences do include Microsoft Defender and Exchange Online Protection to protect against e-mail-based attacks, with E5 licences including the addition of Microsoft Defender for Identity and Endpoint. E3 customers have only the essential levels of protection, while E5 customers have advanced levels of protection and detection of e-mail-borne attacks.

But even so, there are additional layers of your defence-in-depth strategy to harden the e-mail perimeter that can be taken to make it increasingly more difficult for cyber attackers to navigate. The goal is to stop any zero-day attacks that first need to establish a foothold via phishing.

Several technologies exist today to ensure the validity and integrity of e-mails received. The Domain-Based Message Authentication, Reporting and Conformance (DMARC) defines a policy around what should be done with an e-mail where the sending domain appears to be impersonated.

The Sender Policy Framework (SPF) is used within DNS to identify the hostnames and IP addresses of valid e-mail senders for a given domain. Putting these in place helps eliminate the possibility of successful domain impersonation on the part of the cyber attacker.

Sending domains can be validated to ensure phishing e-mails aren’t impersonating a trusted brand, while links and redirects can be intelligently followed to see if they take potential victims to spoofed login pages to Office 365, Outlook and other Microsoft 365 cloud services. Phishing kits used to quickly prop up an entire fake website front-end for harvesting can also be detected.

The use of AI focusing on the relationships and connections between senders and recipients, including the strength or proximity of the relationships, can help identify anomalous e-mails that may be malicious.

A strong defence rests in it being based on shared threat intelligence to ensure the most up-to-date data to increase detections and reduce risk. In addition, the intel gathered through protecting the e-mail perimeter should be shared back to your SIEM to aid in providing a comprehensive view of what’s happening on your network.

The opening of an e-mail can be simulated within a virtual environment, where attachments can be detonated to see whether they perform a malicious action.

Links can be scanned in real-time and blocked from being clicked if deemed malicious.

Uninterrupted business continuity is critical, keep e-mail flowing during planned and unplanned outages with a mailbox continuity solution that provides guaranteed access to live and historic e-mail and attachments from Outlook and Windows, the web and mobile applications – from anywhere on any device.

Solutions like Mimecast offer comprehensive e-mail security and resilience solutions. AI-powered detection blocks all e-mail-based threats, while easy-to-deploy, easy-to-manage complementary solutions reduce risk, cost and complexity.

As a certified managed security service provider (MSSP), J2 Software also provides end-to-end support to bolster its clients’ cyber resilience and provides a solid framework for beneficial cyber insurance policies.