Office 365: A Favorite for Cyberattack Persistence
Bad actors are leveraging legitimate services and tools within Microsoft’s productivity suite to launch cyberattacks on COVID-19 stay-at-home workers, new research finds.
Threat actors are consistently leveraging legitimate services and tools from within Microsoft Office 365 to pilfer sensitive data and launch phishing, ransomware, and other attacks across corporate networks from a persistent position inside the cloud-based suite, new research has found.
Office 365 user account takeover – particularly during the COVID-19 pandemic with so many working from home – is one of the most effective ways for an attacker to gain a foothold in an organization’s network, said Chris Morales, head of security analytics at Vectra AI.
From there, attackers can move laterally to launch attacks, something that researchers observed in 96 percent of the 4 million Office 365 customers sampled between June to August 2020. The company revealed the findings of this research in a 2020 Spotlight Report, released Tuesday.
“We expect this trend to magnify in the months ahead,” Morales said in an email interview with Threatpost.
The report takes a dive into some of the most popular ways that attackers leverage Office 365 services and tools to compromise corporate networks. Indeed, Office 365 presents a wide playing field for attackers; the leading software-as-a-service (SaaS) productivity suite has more than 250 million active users each month, which has made it a historically consistent target for attacks.
Many of those users are currently working from home due to COVID-19 restrictions, often on networks that don’t have the same protections as the corporate cloud. This adds another aspect of accessibility for attackers, Morales said.
Researchers found three key features of the suite that attackers exploit to take over accounts and go on to perform a variety of attacks: OAuth, Power Automate and eDiscovery.
“OAuth is used for establishing a foothold, Power Automate is used for command and control and lateral movement, and eDiscovery is used for reconnaissance and exfiltration,” Morales told Threatpost.
OAuth is an open standard for access authentication used in Office 365 and already has been observed by researchers as a way for attackers to gain access to the cloud-based suite. Third-party applications use the standard to authenticate users by employing Office 365 login services and the user’s associated credentials so that they don’t have “to continuously log into every app every time the user and app requires access,” Morales said.
Unfortunately, this convenience also is a boon for threat actors because it allows an attacker to steal OAuth credentials or access them by convincing a legitimate user to approve a malicious app (via phishing email), he said. This can allow attackers to maintain persistent and undetected access to Office 365 accounts.
Power Automate lets users create custom integrations and automated workflows between Office 365 applications, is enabled by default, and includes connectors to hundreds of third-party applications and services—also giving it appeal for both users and hackers, Morales noted.
It allows users to automate mundane tasks but can also be leveraged by attackers, not only because of its default on status, but also because it allows them to make lateral movements within the app and execute malicious command-and-control behaviors, he said.
“There is no way to turn off individual connectors — it is all or nothing,” Morales told Threatpost. “Attackers can sign up for free trials to get access to premium connectors that do even more.”
Vectra found that 71 percent of customers sampled in their research exhibited suspicious Office 365 Power Automate behaviors.
Meanwhile, Microsoft eDiscovery searches across Office 365 applications and data and exports the results. Once inside Office 365, attackers are using this feature as an internal reconnaissance and data exfiltration tool to find critical data to steal that can be used with malicious intent. Fifty-six percent of customers sampled in Vectra’s research exhibited suspicious Office 365 eDiscovery behaviors, researchers found.
Account Compromise Impact
Once attackers use these features and services to take over Office 365 accounts, there are a number of techniques they use to compromise networks. They can search through emails, chat histories, and files looking for passwords or interesting data to exfiltrate, or set up forwarding rules to get access to a steady stream of email without needing to sign-in again, researchers said.
Threat actors also can leverage the trusted communication channel to send socially engineered phishing emails to employees, customers, or partners. For instance, researchers observed (and helped mitigate) an incident where a medical research unit at a university was targeted with a phishing lure that promoted a free calendar optimization and time-management app.
After one person took the bait and installed the malicious OAuth app, the attackers had complete access to Office 365 and used it to send internal phishing emails, taking advantage of trusted identities and communications to spread further inside the university.
Other attacks that can occur due to Office 365 account takeover include the ability to plant malware or malicious links in documents that many people trust and use; or steal or hold files and data for ransom.
To mitigate these threats, researchers recommend that organizations move away from employing static, prevention-based, policy control-centric or one-off mitigations and move to a more contextual security approach, Morales said.
“These approaches continue to fail,” he told Threatpost. “Security teams must have detailed context that explains how entities utilize their privileges – known as observed privilege – within SaaS applications like Office 365. Just as attackers observe or infer interactions between entities, defenders should think similarly about their adversaries. It is about the usage patterns and behaviors, not the static access.”
- Google Patches Actively-Exploited Zero-Day Bug in Chrome Browser - October 21, 2020
- Google’s Waze Can Allow Hackers to Identify and Track Users - October 20, 2020
- TikTok Launches Bug Bounty Program Amid Security Snafus - October 16, 2020