With credential theft becoming prevalent, through phishing and social engineering, it’s no longer enough to protect from the outside. Attackers may be using internal identities to snoop around and target internal assets, from the inside.
In the face of such attacks, organizations can leverage Advanced Threat Analytics, a solution that would shows and alarms on abnormal user behavior. This is not possible with the traditional security solutions that are in use.
The solution can be easily deployed, without an agent, and without causing any loss of performance. Advanced Threat Analytics can be integrated with existing SIEM tools.
Security teams can rest assured that they don’t need to write any rules for known attack types –Advanced Threat Analytics provides an Advanced Threat Analytics base that is regularly updated for new attack types.
From detecting known malicious attacks to uncovering abnormal activity with machine learning and behavioral analytics, identify advanced persistent threats to your enterprise quickly and take action swiftly with Microsoft Advanced Threat Analytics.
Click here to see a recent blog article: " Phishing Campaign Thwarted in 10 Minutes."
Advanced Threat Analytics Identifies Attacks
Advanced Threat Analytics provides detection for the following various phases of an advanced attack: reconnaissance, credential compromise, lateral movement, privilege escalation, domain dominance and others. These detections are aimed at detecting advanced attacks and insider threats before they cause damage to your organization. The detection of each phase results in several suspicious activities relevant for the phase in question, where each suspicious activity correlates to different flavors of possible attacks. These phases in the kill-chain where Advanced Threat Analytics currently provides detections are highlighted in the image below.
Advanced Threat Analytics provides multiple reconnaissance detections. These detections include:
- Reconnaissance using DNS - DNS information in the target network is often very useful reconnaissance information. DNS information contains a list of all the servers and often all the clients and the mapping to their IP addresses. Viewing DNS information may provide attackers with a detailed view of these entities in your environment allowing attackers to focus their efforts on the relevant entities for the campaign.
- Reconnaissance using directory services enumeration - Detecting reconnaissance for entities (users, groups, etc.) performed using the SAM-remote protocol to run queries against the domain controllers. This reconnaissance method is prevalent in many types of malware seen in real-world attack scenarios.
To provide detection of compromised credentials, Advanced Threat Analytics leverages both machine-learning based behavioral analytics as well as known malicious attacks and technique detection. Using behavioral analytics and machine learning, Advanced Threat Analytics is able to detect suspicious activities such as anomalous logins, abnormal resource access, and abnormal working hours which would point to credential compromise. To protect against compromised credentials, Advanced Threat Analytics detects the following known malicious attacks and techniques:
- Brute force In brute-force attacks, attackers try to guess user credentials by trying multiple users and pairing them with multiple password attempts. The attackers often use complex algorithms or dictionaries to try as many values as a system allows.
- Sensitive account exposed in plain text authentication If high-privileged account credentials are sent in plain text, Advanced Threat Analytics alerts you so that you can update the computer configuration.
- Service exposing accounts in plain text authentication If a service on a computer is sending multiple account credentials in plain text, Advanced Threat Analytics alerts you so that you can update the service configuration.
- Honey Token account suspicious activities Honey Token accounts are dummy accounts set up to trap, identify, and track malicious activity that attempts to use these dummy accounts. Advanced Threat Analytics alerts you to any activities across these Honey Tokens accounts.
- Unusual protocol implementation Authentication requests (Kerberos or NTLM) are usually performed using a normal set of methods and protocols. However, in order to successfully authenticate, the request only has to meet a specific set of requirements. Attackers can implement these protocols with minor deviations from the normal implementation in the environment. These deviations may indicate the presence of an attacker attempting to leverage or successfully leveraging compromised credentials.
- Malicious DAdvanced Threat Analytics Protection Private Information Request DAdvanced Threat Analytics Protection API (DPAPI) is a password-based DAdvanced Threat Analytics protection service. This protection service is used by various applications that stores user’s secrets, such as website passwords and file share credentials. In order to support password-loss scenarios, users can decrypt protected DAdvanced Threat Analytics by using a recovery key which does not involve their password. In a domain environment, attackers may remotely steal the recovery key and use it to decrypt protected DAdvanced Threat Analytics in all the domain joined computers.
- Abnormal Behavior Often in cases of insider threats, as well as advanced attacks, the account credentials may be compromised using social engineering methods or new and not-yet-known methods and techniques. Advanced Threat Analytics is able to detect these types of compromises by analyzing the entity’s behavior and detecting and alerting on abnormalities of the operations performed by the entity.
To provide detection of lateral movement, when users take advantage of credentials that provide access to some resources to gain access resources that they are not meant to have access to, Advanced Threat Analytics leverages both machine-learning based behavioral analytics as well as known malicious attacks and technique detection. Using behavioral analytics and machine learning, Advanced Threat Analytics detects abnormal resource access, abnormal devices used and other indicators that are evidence of lateral movement. In addition, Advanced Threat Analytics is able to detect lateral movement by detecting the techniques used by attackers to perform lateral movement, such as:
- Pass the ticket In pass the ticket attacks, attackers steal a Kerberos ticket from one computer and use it to gain access to another computer by impersonating an entity on your network.
- Pass the hash In pass the hash attacks, attackers steal the NTLM hash of an entity, and use it to authenticate with NTLM and impersonate that entity and gain access to resources on your network.
- Over-pass the hash Over-pass the hash are attacks in which the attacker uses a stolen NTLM hash to authenticate with Kerberos, and obtain a valid Kerberos TGT ticket, which is then used to authenticate as a valid user and gain access to resources on your network.
- Abnormal behavior Lateral movement is a technique often used by attackers, to move between devices and areas in the victim’s network to gain access to privileged credentials or sensitive information of interest to the attacker. Advanced Threat Analytics is able to detect lateral movement by analyzing the behavior of users, devices and their relationship inside the corporate network, and detect on any abnormal access patterns which may indicate a lateral movement performed by an attacker.
Advanced Threat Analytics detects successful and attempted privilege escalation attacks, in which attackers attempt to increase existing privileges and use them multiple times in order to eventually gain full control over the victim’s environment. Advanced Threat Analytics enables privilege escalation detection by combining behavioral analytics to detect anomalous behavior of privileged accounts as well as detecting known and malicious attacks and techniques that are often used to escalate privileges such as:
- MS14-068 exploit (Forged PAC) Forged PAC are attacks in which the attacker plants authorization DAdvanced Threat Analytics in their valid TGT ticket in the form of a forged authorization header that grants them additional permissions that they weren't granted by their organization. In this scenario the attacker leverages previously compromised credentials, or credentials harvested during lateral movement operations.
- MS11-013 exploit (Silver PAC) MS11-013 exploit attacks are an elevation of privilege vulnerability in Kerberos which allows for certain aspects of a Kerberos service ticket to be forged. A malicious user or attacker who successfully exploited this vulnerability could obtain a token with elevated privileges on the Domain Controller. In this scenario the attacker leverages previously compromised credentials, or credentials harvested during lateral movement operations.
Advanced Threat Analytics detects attackers attempting or successfully achieving total control and dominance over the victim’s environment by performing detection over known techniques used by attackers, which include:
- Skeleton key malware In skeleton key attacks, malware is installed on your domain controller that allows attackers to authenticate as any user, while still enabling legitimate users to log on.
- Golden ticket In golden ticket attacks, an attacker steals the KBTGT's credentials, the Kerberos Golden Ticket. That ticket enables the attacker to create a TGT ticket offline, to be used to gain access to resources in the network.
- Remote execution Attackers can attempt to control your network by running code remotely on your domain controller.
- Malicious replication requests In Active Directory (AD) environments replication happens regularly between Domain Controllers. An attacker can spoof AD replication request (sometimes impersonating as a Domain Controller) allowing the attacker to retrieve the DAdvanced Threat Analytics stored in AD, including password hashes, without utilizing more intrusive techniques like Volume Shadow Copy.