Endpoint Security

Endpoint security refers to the strategies and technologies for preventing, containing, mitigating, and remediating threats to endpoints. Endpoint threats can involve external attacks as well as insider threats, which may be either malicious or unintentional in nature. A compromised endpoint can give an attacker a foothold within an environment. From that foothold, the threat actor can launch further attacks on systems to access data and compromise additional endpoints via lateral movement.

This glossary post will explore:

• The challenges of managing and securing endpoints
• Key endpoint attack vectors
• Strategies, technologies, tools, and solutions for implementing endpoint security

What is an Endpoint?

An endpoint is any physical or virtual device or hardware that connects to the corporate environment. An endpoint will either use a TCP/IP (v4 or v6) address or another protocol for networking (wired or wireless). Endpoints are anything from end-user devices, such as desktop PCs, laptops, tablets, and smartphones to servers, medical devices, IoT devices and sensors, industrial control systems (ICS), point-of-sale (PoS) devices, ATMs, printers, network switches, routers, wearables, and much more.

Some endpoint security tools and management technologies can be applied across a diverse variety of endpoints. Other solutions are device or platform-specific (Windows, macOS, Unix, Linux, Android, iOS, Azure, AWS, Google Cloud etc.).

The grid below roughly breaks down the device coverage across some common types of endpoint security solutions. Coverage may vary even between different vendor product offerings within a solution space.

Why is endpoint security important?

According to a 2020 Ponemon/Morphisec Study, 68% of organizations have experienced one or more endpoint attacks over the last two years. Reports also indicate that anti-virus / anti-malware solutions miss an average of 60 percent of attacks! Of the attacks that resulted in a compromised endpoint, 80% were characterized as a zero day threats (for which signature-based antivirus/antimalware is ineffective), but 17% say the threat was known, and 3% are unsure.

The defense-in-depth approach to endpoint security

Traditional AV/anti-malware can only account for part of a multi-layered approach to protecting your endpoint universe. Your endpoint protection should offer layered protection against both internal and external threats, whether you are on-premise, remote, or connecting via the cloud. This is the defense-in-depth approach. Layering your endpoint protection by combining the right endpoint security tools is the best way to eliminate gaps, reduce attack surfaces, and contain threats.

When Do You Need Endpoint Protection?

Endpoint security is one of the broadest cybersecurity categories. It overlaps, integrates with, and complements network security, data security, identity-based security, and application security. Many IT security solutions provide capabilities that cover multiple security categories across endpoint, network, data, and identity-based models.

A corporate IT network is essentially a linkage of endpoints. This means endpoint integrity and security may be prioritized before implementing security at the network layer.

How Do Endpoint Security Tools Work?

Endpoint protection strategies and solutions aim to secure endpoints, whether they are connected to the network or have transient interconnectivity.

Endpoint security software may be:

• Centrally managed via agentless technology
• Installed as client or agent on individual endpoints
• Cloud-based
• Involve a combination of the above implementations

Protection for individual consumers focuses on technologies deployed on the device. Enterprises, on the other hand, should heavily rely on centralized management across the corporate network. This enables the administration of patches, configuration changes, deployment of policy updates, gather logs, and more.

The Differences Between Rules-Based, Signature-based, and Behavior-based Security

Endpoint security strategies often leverage the following three approaches to detect, prevent, and/or mitigate threats:

Pattern matching and signature-based (i.e. traditional antivirus, vulnerability management, etc.): Relies on threat signatures to block known threats. They may also use heuristics to block suspicious code or actions that share similarities with known threats.

Rules-based and enforcement of advanced policies (i.e. privileged access management, endpoint firewalls, encryption, etc.): Applies rules and policies that enforce security best practices, such as least privilege, block lists and allow lists, endpoint firewall rules, and more. The results are predictable based on rule matching processing.

Behavior-based (endpoint detection and response, etc.): Modern endpoint protection may apply advanced behavioral analysis, machine learning, and even some forms of artificial intelligence to identify threats or inappropriate access. These solutions can process information locally or rely on management servers to aggregate information for advanced detection and response.

Strategies for securing endpoints encompass such practices as endpoint hardening, endpoint isolation, endpoint lifecycle and policy management, and more. These are typically signature and rule-based.

There are many colliding factors that put an organization’s universe of endpoints at risk. Examples include evolving cyber threats, complex and diverse endpoint environments, corporate misalignment of security technologies to threats, and an ever-more stretched IT team. Many technologies deployed as part of an endpoint security strategy provide overlapping capabilities in some areas, while leaving gaps in other areas. Overlapping features may also cause incompatibility issues and produce unpredictable results, or degrade endpoint stability.

Let’s explore the 15 most significant challenges and weaknesses present with enterprise endpoints:

1. Lack of visibility into the endpoint estate

Most organizations struggle to understand the scope of their endpoint estate, let alone how to manage it. An organization may have millions of endpoints, and even tens of thousands of undiscovered or unknown endpoints that connect to the network.

2. Overwhelming number and diversity of endpoints

The sheer number and diversity of corporate endpoints presents massive challenges and makes it hard to standardize security. Non-traditional endpoints proliferate. IoT devices, such as sensors, security cameras, healthcare devices, wearables, etc. are pervasive.

Many IT teams struggle to discover and securely onboard legitimate devices at scale. Compounding this issue, IoT and edge computing devices commonly have hardcoded credentials, software or firmware that cannot be updated, and other severe security drawbacks. The devices may use unsafe, or uncommon protocols, and even have built-in backdoors. Because most IoT devices have little computing power, they most likely cannot run security software or host an agent.

Over the last decade, sophisticated attacks, such as Stuxnet, waged on industrial control systems (ICS) has been a growing concern. ICS endpoints were traditionally “air-gapped”, meaning they were isolated from internet-facing networks. These critical infrastructure endpoints are becoming increasingly connected, and thus, exposed to cyber risk. ICS endpoints can include programmable logic controllers (PLCs), supervisory control and data acquisition (SCADA) systems remote terminal units (RTUs), intelligent electronic devices (IEDs), human machine interfaces (HMI), and much more.

The rise of the virtual desktop is another challenge. Virtual desktops share many of the risks as physical desktops, but their dynamic and ephemeral nature adds further complexity.

3. Threat actors

Threat actors continue to learn new tricks and acquire more sophisticated tools with powerful automation capabilities. Modern cybercriminals can easily scale attacks across tens of thousands of endpoints and target across organizations, types of devices, and many other ways.

Attackers can readily acquire toolkits off the dark web, with nation-state caliber tools. These tools can be anything from credential hacking tools and password dictionaries, to advanced, polymorphic malware that incorporates machine learning. In addition, malware is ever-improving at evading antivirus and other security defenses, and blending in with normal processes.

Even phishing, particularly spear phishing, attacks are getting more savvy. Spear phishers may use scanning tools to mine relevant data across vast resources. They may then leverage machine learning to craft phishing messages that target their victim with eerie precision.

While the vectors used by threat actors are too broad to cover here, one worth mentioning is Living off the Land (LotL) attacks. LotL attacks refers to the leveraging of tools or features that preexist within the target’s environment. These are considered fileless attacks because no execution of files (i.e. malware) is needed to perform their illicit activities. LotL attacks essentially turn a system's own native applications and commands against itself. PowerShell scripts, VB scripts, WMI, and PsExec., and Windows script host executables are common culprits in fileless attacks.

Finally, another potential threat that has lurked on endpoints for decades, but gets little attention today, is spyware. While this unwanted software is often merely an annoyance, it can compromise privacy, leak data, and potentially perform other malicious actions.

4. Overprovisioning of privileges/privileged access

Managing privileged access on endpoints is critical because endpoints typically have far more privileged access and default system privileges than necessary. Almost every single cyberattack today involves the use of privileged access.

Default privileges are created and managed differently across different operating systems. Here are a few examples:

• In Windows systems, the Administrator account holds superuser privileges, and each computer has at least one local administrator account. Typically, local admin rights far exceed what most users need. Administrative rights should be removed from most non-IT users, and some endpoints should be configured with only guest account access.
• In Linux and Unix-like systems, the superuser account, called ‘root’, has unrestricted access to all commands, files, directories, and resources. Organizations should eliminate root access and put tight security controls and session auditing around any superuser access. The sudo (“superuser do”) command allows the user to temporarily elevate privileges to root-level without having direct access to the root account and password. This may suffice in some instances, but it introduces its own security and administration issues.
• For macOS endpoints, IT may need to create non-privileged accounts for users (standard users), and manage the local administrator account.

Often, these privileges are persistent (also called “standing privileges”), meaning they are constantly in an active state, ready to be used or exploited. Each instance of excessive privilege risks a privilege-based attack. It also risks a lateral attack. Threat actors exploit privileges to gain an initial foothold, then escalate privileged access to other endpoints and assets across the corporate network. According to research published by IBM, 70% of attacks involve attempts to move laterally.

5. Endpoint vulnerabilities

Endpoint vulnerabilities represent one of the largest security exposures. Most mid-to-large sized organizations will have thousands of software vulnerabilities strewn across their endpoints due to the infeasibility of remediation. Fixing a vulnerability on an endpoint can pose its own risks, such as downtime.

Organizations must assess the risks and benefits of remediation. Indeed, many vulnerabilities are known and present across an enterprise for many years. Of course, the risk increases when there is a known exploit targeting a vulnerability. According to a Ponemon/Morphisec Study, it takes organizations, on average, 97 days to apply, test, and fully deploy patches.

Here are the three main types of software vulnerabilities found on endpoints:

• Firmware vulnerabilities: Firmware is the code/software built into a specific hardware that enables operation of basic device-level functions. On some devices, firmware is loosely considered to be the device’s operating system (iOS for Apple iPhones). Because they enable operation of basic functions, firmware vulnerabilities can be dangerous. Some firmware vulnerabilities essentially give a threat actor root access to a device, completely undermining security features like encryption. Firmware vulnerabilities have become a concern with the proliferation of devices like IoT and smartphones; however, they can also arise in mature vendor products.
• Operating system (OS) vulnerabilities can be errors in code or shortcomings of actual features that may pose a risk, such as permitting unnecessary access. Common operating systems found in enterprise environments include Windows, macOS, Unix, and Linux.
• Application vulnerabilities are bugs or weaknesses in installed applications. Vulnerabilities could include anything from privilege application vulnerability to backdoor root access, to embedded default credentials. Unlike with firmware and OS systems, end users often have the privileges to install unvetted applications (known as shadow IT). This poses an elevated risk.

6. Deficiencies in native OS security and tool sets

As mature as Windows is, many of the OS tool sets are basic and fall far short of the security needed within enterprise environments. Native Windows security components provide only basic protections. These include Windows Defender Antivirus, Windows Firewall (called Windows Defender Firewall in Windows 10), and Local Administrator Password Solution (LAPS). These native technologies are also often disabled within enterprises. On the other hand, Windows BitLocker encryption is often an important part of corporate endpoint security.

Linux has some advanced auditing capabilities, but that's far from enough. The sudo program enables users to elevate system privileges on Unix and Linux systems, without receiving root access. Sudo helps mitigate threats given no alternative. However, sudo itself has many limitations and introduces its own security risks. Put simply, sudo falls woefully short of the enterprise-class privileged access security needed for these environments.

macOS systems have traditionally lagged in security, management, and monitoring tool sets, though security has received a bigger focus from Apple in recent years.

7. Shadow IT

Shadow IT refers to any IT technology within a corporate environment that is not provisioned by IT. Users either install it on the endpoint themselves or access it via the cloud. It can refer to an unauthorized device that the user has connected to the network. When end users self-provision applications or connect unauthorized endpoints to the network, it can introduce a number of security and operational issues. Shadow IT may be incompatible or interfere with other network systems, suffer from dangerous vulnerabilities, have excessive privileges, and may even introduce a backdoor.

Controlling shadow IT has been a decade-long IT challenge. The early months of the coronavirus pandemic saw users scrambling to implement tools that could help them maintain productivity while working from home. This caused shadow IT to spike. In some cases, this included insecure remote access and teleconferencing tools, some of which attackers promptly exploited, with devastating results.

8. Insider threats ​

Insider threats have always presented one of the most significant risks to corporate endpoints and networks. This is because internal users (employees, vendors, contractors, etc.) already have access on the endpoint. These users tend to fly under the radar of many traditional threat detection solutions. An employee could infect their device by inadvertently opening corrupted content from a phishing email. They could also visit a malicious website that initiates a drive-by download of malware. If the insider is ill-bent and possesses high levels of privileged access, they can easily inflict catastrophic damage. Insider threats, when they happen, tend to be the most harmful.

The 2019 Human Factor Report by Proofpoint found that more than 99% of threats observed required human interaction to infect user devices.

9. Immature password lifecycle management and authentication

Endpoints – desktops, servers, mobile devices, or IoT - almost always have credentials, and many of these are privileged. Privileged credentials allow the user (human or machine) to perform powerful functionalities or access sensitive devices or applications. According to Forrester Research, over 80% of breaches today involve privileged credentials. Verizon’s 2020 Data Breach Investigations Report (DBIR) found that threat actors are increasingly prioritizing acquisition of credentials over deployment of malware. After all, with the right credentials, an attacker obtains instant access to the network. Unfortunately, poor password management practices are commonplace within enterprise environments. Here are just a few:

• Credentials are manually managed, resulting in poor practices (short passwords, uncomplex passwords, and password reuse). The shorter and less complex the password, the easier it is for automated tools to crack it. If a duplicated password becomes compromised, a threat actor can use those stolen credentials to compromise any endpoints that share the same credentials. Personal devices typically have much lower cyber defenses than enterprise endpoints. When employees repeat the same passwords across their personal endpoints and work environment, it poses a considerable risk to the corporate network.
• Default and hardcoded passwords may be present in IoT, legacy applications, and various networked devices. Attackers can also find these embedded within compiled code. If the same default password is reused across a device model, that information can be used to compromise all similar devices.
• Credentials are often shared between users, particularly to perform IT administration tasks. This both heightens the risk of password exposure and makes it almost impossible to trace endpoint activity to a single user. A clear session activity trail is a requirement for many common compliance and audit initiatives.

10. Inadequate encryption

Encryption cracking tools are getting increasingly powerful. Yesterday’s encryption standards can’t always withstand modern attacks. Without proper encryption, an unauthorized user or threat actor can easily read and compromise data should they gain access.

11. Lost or stolen endpoints

In the years of mainframes and desktop computing devices, stolen devices were a rarity. However, once devices went mobile, the number of lost or stolen enterprise devices soon skyrocketed. Stolen or lost devices can create security and compliance risks, especially if the information or access provided by the device is sensitive. The potential for security, legal, and brand fallout increases substantially if the device was improperly secured.

12. Insecure remote access pathways

A growing mobile and remote workforce means endpoints commonly operate outside the network perimeter. These endpoints often connect remotely via home networks and unmanaged WiFi. RDP, VNC, and SSH protocols can provide basic remote access. These tools are often stretched beyond their appropriate use cases, which poses serious security and operational ramifications.

RDP by itself is insecure. Organizations using RDP should wrap additional security and connectivity solutions around it to prevent the exposure of remote access sessions.

An employee should never use VPNs to connect to corporate networks from their personal devices (BYOD). Vendor endpoints and internal endpoints that need to perform privileged access remotely require more robust technologies. VPNs can’t granularly control, monitor, or report on privileged access; only access in general.

Attackers sometimes plant remote access technologies on targeted endpoints (such as a vendor or IT service provider support systems). The attackers then leverage the remote access on those devices as a jumping off point for attacks on the vendor/IT service provider’s customers.

Almost all cyberattacks leverage remote access as a vector, so it’s a critical security piece to get right.

13. The use of personal devices (BYOD) 

The use of personal devices for work introduces risks and challenges. A personal device may lack many of the endpoint protections and hardening that are applied to corporate-provisioned devices. It’s also possible that family members or housemates share the device and have accounts on it too. Any of these users’ activities could imperil the corporate network. BYOD and VPN should never mix, yet, organizations frequently allow protocol tunneling via these methods.

Another complication that occurs with bring your own device deployments is what to do when a device is lost or stolen. The device owner may object to a remote data wipe because it threatens the potential loss of person content. This may have high personal value (such as family pictures or videos). However, if that device has any potential to be in the wrong hands, it's in the company's best interest to instantly disable it.

14. Vendor access/vendor endpoints 

Vendor access represents perhaps the weakest cybersecurity link for most organizations. According to research published by Opus & Ponemon, the average organization shares confidential and sensitive information with approximately 583 third parties. Threat research published by BeyondTrust found that organizations, have, on average, 182 vendors logging into their systems every week. That adds up to a lot of potential security “weakest links!”

Opus and Ponemon reported that 59% of companies incurred a breach due to a vendor, while the BeyondTrust report put that number at 58%. Why are so many organizations incurring breaches as a result of vendor endpoints and access?

VPNs and many other commonly used technologies don’t provide the granular access controls needed. Many organizations are unable to identify the vendor endpoints accessing their systems, let alone monitor or control what is occurring during the sessions. Do the vendor’s own security standards meet those of your enterprise? How do you enforce and validate that? Do some of a vendor’s employees use personal devices to access your network? What happens if these employees are terminated or leave—could their access remain as an orphan account? The risks arising from vendor endpoints are massive.

Since vendors play such an important role for most organizations, getting the vendor access security piece correct is imperative.

15. Misconfigured endpoints

These may not only hamper operations, but may also introduce attack vectors. Poor configuration settings could involve open ports, outdated exceptions, insecure protocols allowed, etc.

No single strategy or technology is by itself sufficient to protect your entire, heterogeneous endpoint universe from all threats. Some strategies (endpoint hardening) and technologies (privilege management) may be effective security controls against the vast majority of threats on any type of endpoint. Other strategies (remote wiping, anti-theft protection, etc.) and technologies (antivirus) may only work, or be applicable, for specific types of endpoints.

Every enterprise should adopt the following endpoint security best practices:

1. Define, communicate, and consistently apply your endpoint protection policy

Your endpoint protection policy is a living, evolving document that should be part of your overarching IT security policy. It should programmatically define best practices and how they should be applied. The policy should define strategies, technologies, and data ownership. An example of this is BYOD. If your organization allows BYOD, the policy should clearly define which data is owned by the company and which belongs to the end user. The policy should detail what should happen in case of a lost or stolen device.

Any personal device used for corporate access should also have some form of mobile device management installed on it. MDM can enforce separation of personal and corporate resources, and provide other security functionalities. Your policy should also clarify the protocols surrounding vendor controls. It should define the security controls you will put around vendor access. IT should also define the security controls you expect the vendor to implement around their own endpoints. Finally, ensure there is a method in place to measure, test, and audit all of these policies to ensure they are functioning as desired.

2. Discover and onboard, or deny access to, endpoints

The next step involves finding and inventorying every endpoint that connects to your network and onboarding them. This implies applying your organization’s security policies to the endpoints after they are discovered. Discovery and onboarding should be a continuous process. In some cases, this step will occur proactively, such as when IT is provisioning a device for a user or deploying a server. In other cases, this will occur reactively, such as when a new, unknown device touches the network. Devices should also be properly registered and monitored for issues, necessary updates, and to reconfirm overall health status.

3. Endpoint hardening 

Endpoint hardening involves removing or disabling unnecessary programs, access, embedded or other features to condense the attack surface. It is a continuous process performed throughout the lifecycle of technology—from initial installation, through configuration, maintenance, and support, to end-of-life decommissioning. Endpoint hardening practices vary by endpoint, with more complex and powerful endpoints generally requiring more steps. Endpoints should always be hardened before being connected to the internet or corporate networks. Application and Operating system hardening are two of the most important aspects of endpoint hardening.

OS hardening can entail the following:

• Removing unnecessary drivers, file sharing, libraries, software, services, and functionality
• Encrypting local storage
• Tightening registry and other systems permissions
• Implementing privileged user controls

Application hardening can entail:

• Restricting app-to-app communications
• Removing or turning off unneeded features
• Eliminating embedded credentials and replacing them with API calls

Endpoint hardening is really a subset of actions aligned with the principle of least privilege (PoLP), which is described next.

4. Enforce least privilege

Privileged accounts and access should either be outright eliminated, or restricted to the lowest level required for endpoints, system processes, software, applications, and users. The 2021 Microsoft Vulnerability Report revealed that eliminating local admin rights on endpoints would have mitigated 56% of the Critical Microsoft vulnerabilities in 2020. BeyondTrust CTO/CISO Morey Haber published vulnerability research showing that similar reductions could also be achieved by enforcing least privilege across third-party applications.

Since most malware requires privilege to execute, least privilege can prevent attacks from compromising an endpoint in the first place. If an endpoint has already been successfully breached, the attack is contained. Malware and cybercriminals typically need privileged access to move laterally around the environment. Reducing privilege to the minimum level means that, even if attackers gain a foothold, they are essentially marooned.

Further, users should only be able to log into privileged accounts (particularly for Administrator, Domain Admin, or Root) when absolutely necessary, and for only the finite moments when privileged access is required (just-in-time privileged access). This allows you to condense the threat window by limiting the duration of time when privileges are actually available for use and exploit.

The ability for least privilege to work across three dimensions makes least privilege arguably the most important and powerful security best practice. It can prevent attacks from landing and executing, limit and prevent movement, and restrict privilege availability based on time. Least privilege can protect against both known and unknown threats, whether they originate internally or externally.

5. Apply encryption to protect data stored on the device and in transit

To minimize the risk of data leaks, the strength of encryption should at least meet what is required by applicable regulatory standards. More sensitive data (for instance, health data or superuser credentials), may necessitate higher levels of encryption than less sensitive data. Even if a device is stolen or an email is intercepted, if robust encryption is in place, the data will be unreadable. The theft or loss of adequately encrypted devices is not considered a reportable data breach under many regulatory and compliance initiatives. This includes the HIPAA Breach Notification Rule.

6. Identify, prioritize, and remediate vulnerabilities

The vulnerability management process involves scanning for and assessing the impact of known vulnerabilities (i.e. CVEs). These are compared against the operating system, firmware, custom software, and commercial applications that are running on your endpoints. Your organization should have a mature patch management program. It also needs a strong policy that dictates desired software configurations and how any changes should occur, if needed. Also ensure devices are running supported versions of firmware, software, and applications (etc.).

In some cases, risk-benefit analysis may dictate that no remediation action be taken. This often happens when a vulnerability poses a low risk, but the fix for it may disrupt many endpoints or impact mission-critical systems. Vulnerability remediation can entail applying a patch, performing a configuration change, and eliminating embedded or default credentials (etc.). Aside from patch management, vulnerability management may also involve pen testing and threat hunting. Pen testing typically involves third-party security professionals and tools probing for vulnerabilities across your endpoints that can be exploited.

In the absence of a working exploit, a vulnerability by itself is just a theoretical risk. However, once an exploit has been architected for a vulnerability, it becomes a threat. Threat hunting refers to the process of proactively seeking out threats by looking for indicators of compromise (IoCs). When threat hunting, you assume that an attack has already occurred and a threat actor is active in your environment.

7. Implement threat detection, prevention, and mitigation

Many viruses, ransomware, and other malware threats are already known and documented. This knowledge should be applied to block, quarantined, remove, or mitigate these threats. Signature-based threat protection may only protect your environment against up to 40% of malware threats. However, it remains an important line of defense, as well as a requirement for all regulatory compliance mandates.

8. Protect devices from physical threats

This entails such security controls as:

• Implementing anti-theft technology on the devices (i.e. smartphones, PCs)
• Enabling GPS location awareness, providing security cables for laptops
• Restricting physical access to where the devices are used or stored.

For instance, you may want to ensure laptops are locked in a secure cabinet in an area under constant human and/or video surveillance.

9. Secure vendor endpoints and remote access

Organizations should strive to achieve the same security best practices they uphold within their perimeter beyond their perimeter. This means ensuring just-in-time protocols. Only the right identity using the right endpoint (which has been properly hardened) should have access to the right resources within the right context. To ensure this with the vendor account/endpoint, you need to apply basic identity management and privileged identity controls. Important controls include:

• Enforcing least privilege
• Applying multi-factor authentication
• Rotating passwords and using one-time passwords (OTPs)
• Implementing session management and monitoring to audit and control all vendor/remote access-initiated session activity.

​10. Integrate your endpoint security technologies with each other and across the rest of your security and IT stack

When evaluating any new IT security tool, there are some important questions you should ask yourself first. How does this tool integrate with the rest of my security ecosystem? Does it create synergies, or does it create more complexity and administration issues? Any new tool/solutions should naturally fit within, and become an integrated part of, your IT and IS environment. The more holistic your picture of risk, the more rapidly you can orchestrate a pinpoint response to prevent, mitigate, or remediate attacks.

With the right mix of endpoint security tools, strategies, and technologies tailored to their environment, organizations can realize at least several distinct benefits:

• Improve security: This is the most obvious benefit. Enforcing least privilege and managing vulnerabilities drastically reduces the threat surface. Applying JIT access controls condenses threat windows. Employing signature-based tools keeps endpoints safe from known threats. Implementing this layered strategy drastically drives down the risk of security incidents and data breaches that may arise from external or internal threats.

• Enhance endpoint performance: Eliminating superfluous privileges and hardening devices translates into fewer misconfigurations, incompatibilities, security incidents, and other issues that may cause disruption. Preventing malware infection or endpoint compromise by an unauthorized user also protects against endpoint instability.

• Simplify compliance and auditability: Restricting features in accordance with least privilege and hardening practices reduces endpoint complexity. This also makes it easier to perform auditing activities and prove compliance. Implementing security controls like session monitoring and management and vulnerability assessments may also provide instant visibility and reporting necessary for compliance efforts. The more tightly an endpoint system is integrated, and the better the visibility across the endpoint universe, the more straightforward the path to compliance.

• Enable the enterprise: A strong endpoint cybersecurity posture enables the enterprise in a number of ways. With the right tools, IT can securely enable more types of endpoints and a heterogeneous infrastructure. It can also empower you to pursue business-enabling changes for the environment, including the roll-out of new technologies. For instance, organizations with robust endpoint protection were well-prepared for the seismic shift to remote work during coronavirus pandemic. Increasingly, cybersecurity posture is also assessed before entering a partnership with another company. Cybersecurity posture is also evaluated when companies undergo an acquisition process. Acquiring companies are increasingly sensitive to negative security surprises.

Here are the core endpoint security solutions, ordered by priority:

1. Antivirus / antimalware

Antivirus and antimalware solutions are the most traditional and well-recognized forms of endpoint protection. AV / antimalware typically works by scanning traffic to a device and/or the device’s content for patterns that match a database of virus signatures. While many IT thought leaders have touted the demise of AV, it remains widely deployed.

Basic antivirus software provides detection and protection against known threat signatures. Today, most AV and anti-malware solutions have evolved to incorporate some advanced protection methodologies that incorporate heuristics, behavioral analysis, and even machine learning. Enterprise AV / antimalware solutions are typically installed on a wide array of devices, including servers, desktops and laptops, gateways, and more. These solutions, which may include “next-generation AV”, are centrally managed from the corporate network in most instances. Some devices, such as smartphones, require specialized AV solutions that work at the device-level only.

Many endpoints, such as IoT, lack the computing necessary to install and run AV. As a side-note, cloud-based AV deployments may be of some use in these cases. Other well-known drawbacks of solutions in this class include:

• Device or network performance degradation when the solution is running
• High numbers of missed threats
• Numerous false positives and security alerts that hinder IT’s ability to respond to what is important

‘Next-generation’ AV solutions tend to avoid some of these issues by leveraging cloud processing and foregoing the use of bloated agents. Some solutions also include sandboxing capabilities. A sandbox essentially quarantines an unknown/suspicious program and runs it in isolation to see if malicious characteristics present themselves.

2. Privileged Access Management (PAM) solutions

PAM solutions manage privileges for users (human and machine), endpoints, systems, applications, and processes. They also monitor and record activity that occurs during privileged sessions. PAM solutions can be deployed on-premise, in the cloud, or hybrid and are a key requirement of zero trust initiatives. Complete PAM platforms are comprised of the following three functional areas:

• Privileged Credential/Password Management solutions (also called Privileged Account & Session Management (PASM)) enable automated discovery and onboarding of all privileged accounts, secure access to privileged credentials and secrets, and auditing of all privileged activities. By enforcing password security best practices across all types of credentials (passwords, secrets, SSH keys, etc.), Privileged credential management solutions eliminate or mitigate threats, such as password reuse, embedded passwords and DevOps secrets, pass-the-hash (PtH), and stolen credentials. Session monitoring and management capabilities empower organizations to audit all activity and pause or terminate suspicious sessions.
• Endpoint Privilege Management solutions are also called Privilege Elevation & Delegation (PEDM). They combine least privilege management and advanced application control capabilities to condense the endpoint attack surface. These solutions can protect against sophisticated attacks from fileless malware/living off the land (LoTL) attacks, ransomware, and zero day exploits. Some can also enable passwordless administration (the ability to perform administrative functions on an endpoint without privileged credentials). Endpoint privilege management can be applied to Windows, Mac, Unix, Linux systems, network devices, IoT, and more.
• Secure Remote Access solutions apply least privilege and robust audit controls to all remote access sessions required by employees, vendors, and service desks. The granular privilege control and specialized security capabilities of this software far transcends that of VPNs, which it often replaces. Vendor credentials can be securely managed (rotated, encrypted, etc.) from a centralized vault and injected when needed. The vendor never sees them. These PAM solutions also provide granular auditing.

Together, these PAM functionalities can help implement a zero trust endpoint security approach

3. Endpoint protection platforms (EPP) 

Endpoint protection platforms are generally comprised of a multilayer set of security technologies managed through a centralized console. These solutions aim to protect against multiple threat vectors across an enterprise’s endpoint estate. Many EPPs leverage the cloud to perform analysis, eliminating any processing hit to endpoints. Endpoint protection platforms can be delivered as SaaS and managed remotely. They can also be installed directly on devices that have central management software hosted on a server.

EPP solutions typically perform scanning of files via an advanced AV/antimalware engine that protects against signature-based attacks. This engine also incorporates behavioral analysis to expand threat protection. EPPs usually incorporate endpoint firewalls that help tightly control network traffic through specific ports on the individual endpoints. These solutions may also incorporate basic application control (whitelisting, blacklisting), sandboxing, and machine learning capabilities.

4. Endpoint management

Endpoint management encompasses a range of solutions and processes around managing the lifecycle of endpoints. Endpoint management solutions may be capable of centrally discovering, onboarding, registering, provisioning, updating, monitoring, and troubleshooting endpoint devices. Solutions in this category include mobile device management (MDM), enterprise mobility management (EMM), unified endpoint management (which may integrate a number of capabilities, including MDM, EMM, etc.). Point solutions, like MDM and EMM, work across certain types of mobile devices. UEM solutions, on the other hand, strive to wrap lifecycle management capabilities around a diverse array of devices. UEM solutions may cover servers, desktops, mobile devices, IoT, and more.

Endpoint management solutions play a key role in optimizing device performance, ensuring correct configurations, and establishing a consistent baseline of device hygiene. These solutions also allow for remote management. They can even wrap security controls (such as remote data wipe, anti-tampering, and geo-tracking) around devices. This provides protection in the event that a device is lost or stolen.

5. Endpoint detection and remediation (EDR) solutions 

Endpoint detection and remediation solutions continuously inspect files and applications that enter a device to extend protection beyond simple signature-based threats. This includes protecting against some types of ransomware, zero-day threats, fileless malware, and more advanced attacks.

In contrast to EPP solutions, EDR solutions can provide intrusion detection and a deeper level of advanced threat analysis and forensics. Thus, EDR tools play a bigger role in containing and responding to a security incident or breach event once it has already occurred. However, one of the drawbacks of EDR solutions can be a high number of false positives and security alerts.

Increasingly, IT service providers are offering managed detection and remediation (MDR), which provides the customer with capabilities backed by a team of analysts. More recently, EDR has been evolving into XDR. XDR solutions incorporate data sources beyond the endpoint, such as traffic analysis, so provide a more holistic and context-informed approach than EDR.

6. Encryption

Encryption is the encoding of data so that it is unreadable and unusable without the correct decryption key. As encryption cracking tools have become more powerful, encryption methodologies have had to evolve to ensure data security. Encryption capabilities are important components of many endpoint security solutions (including PAM, DLP, and endpoint firewalls). They are also important native solutions on device firmware and software. Endpoint encryption software typically either protects individual files, or is applied across the entire hard drive, known as full disk encryption.

7. Application control solutions 

Application control solutions secure endpoints, most commonly end-user devices and servers, by preventing them from executing unauthorized applications. Traditionally, these solutions make use of allow lists (whitelists), block lists (blacklists), and greylists. Greylisted applications are applications that have been identified, but have not been added to allow or block lists. Special security rules may be put in place so that greylisted applications can run in some circumstances, as needed.

Application control solutions may also be able to enforce granular control over application usage. For instance, application control could allow a specific user/endpoint to execute a certain function of an application. At the same time, it can disable or block the execution of other application functions for that user/endpoint. Cloud-based reputation services may also be leveraged to determine whether an application is safe to run.

Application control can be sold as a point solution, or be included in other platforms, such as endpoint privilege management, or next-generation firewalls (NGFWs). Integration within these other platforms often allows for natural augmentation of the capabilities that provide advanced application protection. Application control solutions must be finely calibrated. They need to enable the seamless use of authorized applications for legitimate use cases while still preventing against malware, ransomware, and other threats.

8. Patch management solutions

These help automate the process of downloading and applying patches to software. In some instances, your IT team may need to manually apply the patches. However, automating the patching process as much as possible helps organizations scale to meet the thousands of vulnerabilities across their endpoint estate. When dealing with a dangerous vulnerability for which an exploit already exists in the wild, time is of the essence, so automation is key. Patch management is a core part of vulnerability management and is often tightly integrated with enterprise vulnerability management solutions.

9. Vulnerability management (VM) solutions 

Vulnerability management solutions provide a proactive approach to discovering, analyzing, and remediating vulnerabilities. Drawing from vulnerability databases (CVE, etc.) these solutions perform vulnerability scans that deliver a snapshot of the vulnerabilities found across the endpoint environment. VM solutions put context around the risk, weighing that against the mitigation options for the vulnerability. This helps the organization evaluate the scope and level of risk posed by the threat.

Vulnerability scans generally only represent a moment-in-time snapshot, though some scans that are low-impact on resources may run continuously. Scans often need to run on sensitive resources that require privileged access. These are called credentialed scans as they require injection of credentials to be run. An enterprise privileged credential management solution should be to inject the credentials into the VM solution. This will accommodate the highly sensitive and privileged nature of such scans. Credentialed scans tend to uncover more dangerous threats than credential scans.

Vulnerability scanning can be highly automated. Some VM solutions can scale across the entire IT infrastructure. However, human input is typically required to determine and initiate the best course of action. Of course, most vulnerability management solutions are dependent on databases of known vulnerabilities (signatures) and provide limited, if any, protection against zero-day threats.

A good vulnerability management tool should be continuously updated with the latest vulnerability signatures.

10. Web filtering solutions

Web filtering solutions (URL filtering, etc.) put controls around the sites that users can access via their browsers. This helps to mitigate the human element of threats. Organizations can block content by category and/or blacklist certain URLs. Employees reaching a blocked page will typically see a message informing them why their access to the request page has been restricted. Web filtering solutions can secure endpoints by preventing them from visiting malicious sites that pose a risk of infection. Web filtering can also block content that a company deems inappropriate. In some settings, such as k-12 schools, this is a compliance requirement.

11. Data Loss Prevention (DLP) solutions 

DLP solutions aim to prevent leakage or unauthorized transmission of data. DLP is an important endpoint data protection technology. The solutions analyze context and content to help ensure only the right individual or systems can access the data. Some use cases for DLP include preventing a user from:

• Uploading or downloading unauthorized information from a USB
• Accessing a sensitive file attachment in an email when using a personally-owned device
• Emailing a sensitive document (such as containing financial results) to an individual who is not authorized to receive it
• Inappropriately sharing sensitive content with a teammate or other individual over a messaging app—even if the organization authorized the app.

An organization can deploy DLP across many types of endpoints and systems, including end-user devices, servers, and gateways. These solutions leverage encryption, rules-based structure, and database fingerprinting to make fast, accurate decisions about the legitimacy of data access requests. DLP solutions can also send alerts of inappropriate data access requests to the appropriate IT or compliance personnel.

12. Penetration (Pen) Testing solutions

Pen testing allows security teams and researchers to simulate attacks against an environment and probe for potential vulnerabilities. This process yields information useful for hardening the organization’s defenses. Pen testing is a highly proactive methodology that is part of an overall vulnerability management program. Some vulnerability management solutions may have built-in pen testing features. Pen testing is a useful way to find potential threats that other tools may miss. The practice can also contribute a better understanding of how potential changes to IT infrastructure and endpoint configurations could create new attack pathways.

13. Endpoint firewalls/host-based firewalls

Endpoint firewalls are software that runs on endpoints. This is in contrast to the types of firewalls (stateful, packet filtering, next-generation, etc.) more closely associated with network security. Endpoint firewalls are sometimes referred to as “personal firewalls” and they may be installed on and protect desktops, laptops, and servers. Endpoint firewalls inspect traffic, apply rules, and perform behavioral monitoring. This allows them to protect the endpoint from malware and attacks that originate either within or outside the corporate network. A web application firewall (WAF) is a subtype of endpoint firewalls. WAFs are used to secure endpoint-hosted applications and web services, such as WordPress, from SQL injections and other malicious attacks.

​14. Anti-spyware 

Anti-spyware detects, prevents, and removes unwanted software, including adware. While organizations tend to purchase anti-spyware as an add-on module to their anti-virus solutions, standalone software also exists, but is generally consumer-grade.

Evolving to Modern Endpoint Protection

Today, protection of endpoints is necessary to prevent, contain, mitigate, and remediate external and internal threats. Solutions and strategies must scale to meet the growing diversity of devices whether on-premise or remote, employee, or vendor. Most importantly, you should build your endpoint security technology stack with solutions that communicate and collaborate with your broader IT and security ecosystem.

Picking up Where Microsoft Leaves off with Modern Management, Endpoint & Application Security (blog)

How Trusted Application Protection Builds on Application Control & Endpoint Privilege Management (blog)

PAM Buyers Guide & Checklist (paper with buying template)

Passwordless Administration Explained (blog)

Enable & Secure Your Remote Workforce (quick guide)

Just-In-Time Privileged Access Management (quick guide)

5 Ways to Optimize Your Endpoint Protection Strategy With BeyondTrust & McAfee (2-pager)

Complete Privileged Access Management Solutions (web page)