Privilege, in an information technology context, can be defined as the authority a given account or process has within a computing system or network. Privilege provides the authorization to override, or bypass, certain security restraints, and may include permissions to perform such actions as shutting down systems, loading device drivers, configuring networks or systems, provisioning and configuring accounts and cloud instances, etc.

 Privileges serve an important operational purpose by enabling users, applications, and other system processes elevated rights to access certain resources and complete work-related tasks. At the same time, the potential for misuse or abuse of privilege by insiders or outside attackers presents organizations with a formidable security risk.

 Privileges for various user accounts and processes are built into operating systems, file systems, applications, databases, hypervisors, cloud management platforms, etc. Privileges can be also assigned by certain types of privileged users, such as by a system or network administrator.

 Depending on the system, some privilege assignment, or delegation, to people may be based on attributes that are role-based, such as business unit, (e.g., marketing, HR, or IT) as well as a variety of other parameters (e.g., seniority, time of day, special circumstance, etc.).

What are privileged accounts?

 In a least privilege environment, most users are operating with non-privileged accounts 90-100% of the time. Non-privileged accounts, also called least privileged accounts (LUA) general consist of the following two types:

• Standard user accountshave a limited set of privileges, such as for internet browsing, accessing certain types of applications (e.g., MS Office, etc.), and for accessing a limited array of resources, which is often defined by role-based access policies.
• Guest user accountspossess fewer privileges than standard user accounts, as they are usually restricted to just basic application access and internet browsing.

 A privileged account is considered to be any account that provides access and privileges beyond those of non-privileged accounts. A privileged user is any user currently leveraging privileged access, such as through a privileged account. Because of their elevated capabilities and access, privileged users/privileged accounts pose considerably larger risks than non-privileged accounts / non-privileged users.

 Special types of privileged accounts, known as superuser accounts, are primarily used for administration by specialized IT employees and provide virtually unrestrained power to execute commands and make system changes. Superuser accounts are typically known as “Root” in Unix/Linux and “Administrator” in Windows systems.

 Superuser account privileges can provide unrestricted access to files, directories, and resources with full read / write / execute privileges, and the power to render systemic changes across a network, such as creating or installing files or software, modifying files and settings, and deleting users and data. Superusers may even grant and revoke any permissions for other users. If misused, either in error (such as accidentally deleting an important file or mistyping a powerful command) or with malicious intent, these highly privileged accounts can easily wreak catastrophic damage across a system—or even the entire enterprise.

 In Windows systems, each Windows computer has at least one administrator account. The Administrator account allows the user to perform such activities as installing software and changing local configurations and settings.

 Mac OS X, on the other hand is Unix-like, but unlike Unix and Linux, is rarely deployed as a server. Users of Mac endpoints may run with root access as a default. However, as a best security practice, a non-privileged account should be created and used for routine computing to limit the likelihood and scope of privileged threats.

• Local administrative accounts

Non-personal accounts providing administrative access to the local host or instance only.

• Domain administrative accounts

Privileged administrative access across all workstations and servers within the domain.

• Break glass (also called emergency or firecall) accounts

Unprivileged users with administrative access to secure systems in the case of an emergency.

• Service accounts

Privileged local or domain accounts that are used by an application or service to interact with the operating system.

• Active Directory or domain service accounts

Enable password changes to accounts, etc.

• Application accounts

Used by applications to access databases, run batch jobs or scripts, or provide access to other applications.

While most non-IT users should, as a best practice, only have standard user account access, some IT employees may possess multiple accounts, logging in as a standard user to perform routine tasks, while logging into a superuser account to perform administrative activities.

Because administrative accounts possess more privileges, and thus, pose a heightened risk if misused or abused compared to standard user accounts, a PAM best practice is to only use these administrator accounts when absolutely necessary, and for the shortest time needed.

What Are Privileged Credentials?

Privileged credentials (also called privileged passwords) are a subset of credentials that provide elevated access and permissions across accounts, applications, and systems. Privileged passwords can be associated with human, application, service accounts, and more. SSH keys are one type of privileged credential used across enterprises to access servers and open pathways to highly sensitive assets.

Sometimes, especially across DevOps environments, privileged credentials are referred to as “secrets.”

Privileged account passwords are often referred to as “the keys to the IT kingdom,” as, in the case of superuser passwords, they can provide the authenticated user with almost limitless privileged access rights across an organization’s most critical systems and data. With so much power inherent of these privileges, they are ripe for abuse by insiders, and are highly coveted by hackers. Forrester Research estimates that 80% of security breaches involve privileged credentials.

Privileged Risks & Privileged Threats – Why PAM is Needed

Some of the top privilege-related risks and challenges include:

Lack of visibility and awareness of of privileged users, accounts, assets, and credentials: Long-forgotten privileged accounts are commonly sprawled across organizations. These accounts may number in the millions, and provide dangerous backdoors for attackers, including, in many instances, former employees who have left the company but retain access.

Over-provisioning of privileges: If privileged access controls are overly restrictive, they can disrupt user workflows, causing frustration and hindering productivity. Since end users rarely complain about possessing too many privileges, IT admins traditionally provision end users with broad sets of privileges. Additionally, an employee’s role is often fluid and can evolve such that they accumulate new responsibilities and corresponding privileges—while still retaining privileges that they no longer use or require.

All of this privilege excess adds up to a bloated attack surface. Routine computing for employees on personal PC users might entail internet browsing, watching streaming video, use of MS Office and other basic applications, including SaaS (e.g., Salesforce.com, GoogleDocs, etc.). In the case of Windows PCs, users often log in with administrative account privileges—far broader than what is needed. These excessive privileges massively increase the risk that malware or hackers may steal passwords or install malicious code that could be delivered via web surfing or email attachments. The malware or hacker could then leverage the entire set of privileges of the account, accessing data of the infected computer, and even launching an attack against other networked computers or servers.

Shared accounts and passwords: IT teams commonly share root, Windows Administrator, and many other privileged credentials for convenience so workloads and duties can be seamlessly shared as needed. However, with multiple people sharing an account password, it may be impossible to tie actions performed with an account to a single individual. This creates security, auditability, and compliance issues.

Hard-coded / embedded credentials: Privileged credentials are needed to facilitate authentication for app-to-app (A2A) and application-to-database (A2D) communications and access. Applications, systems, network devices, and IoT devices, are commonly shipped—and often deployed—with embedded, default credentials that are easily guessable and pose substantial risk. Additionally, employees will often hardcode secrets in plain text—such as within a script, code, or a file, so it is easily accessible when they need it.

Manual and/or decentralized credential management: Privilege security controls are often immature. Privileged accounts and credentials may be managed differently across various organizational silos, leading to inconsistent enforcement of best practices. Human privilege management processes cannot possibly scale in most IT environments where thousands—or even millions—of privileged accounts, credentials, and assets can exist. With so many systems and accounts to manage, humans invariably take shortcuts, such as re-using credentials across multiple accounts and assets. One compromised account can therefore jeopardize the security of other accounts sharing the same credentials.

Lack of visibility into application and service account privileges:

Applications and service accounts often automatically execute privileged processes to perform actions, as well as to communicate with other applications, services, resources, etc. Applications and service accounts frequently possess excessive privileged access rights by default, and also suffer from other serious security deficiencies.

Siloed identity management tools and processes: Modern IT environments typically run across multiple platforms (e.g., Windows, Mac, Unix, Linux, etc.)—each separately maintained and managed. This practice equates to inconsistent administration for IT, added complexity for end users, and increased cyber risk.

IoT, DevOps, and cloud environments present new privileged threat vectors and privilege management use cases:

Cloud and virtualization administrator consoles (as with AWS, Office 365, etc.) provide nearly boundless superuser capabilities, enabling users to rapidly provision, configure, and delete servers at massive scale. Within these consoles, users can effortlessly spin-up and manage thousands of virtual machines (each with its own set of privileges and privileged accounts). Organizations need the right privileged security controls in place to onboard and manage all of these newly created privileged accounts and credentials at massive scale.

DevOps environments—with their emphasis on speed, cloud deployments, and automation—present many privilege management challenges and risks. Organizations often lack visibility into privileges and other risks posed by containers and other new tools. Inadequate secrets management, embedded passwords, and excessive privilege provisioning are just a few privilege risks rampant across typical DevOps deployments.

IoT devices are now pervasive across enterprises. Many IT teams struggle to discover and securely onboard legitimate devices at scale. Compounding this issue, IoT devices commonly have severe security drawbacks, such as hardcoded, default passwords and the inability to harden software or update firmware.

Privileged Threat Vectors—External & Internal

Hackers, malware, partners, insiders gone rogue, and simple user errors—especially in the case of superuser accounts—comprise the most common privileged threat vectors.

External hackers covet privileged accounts and credentials, knowing that, once obtained, they provide a fast track to an organization’s most critical systems and sensitive data. With privileged credentials in hand, a hacker essentially becomes an “insider”—and that’s a dangerous scenario, as they can easily erase their tracks to avoid detection while they traverse the compromised IT environment.

Hackers often gain an initial foothold through a low-level exploit, such as through a phishing attack on a standard user account, and then skulk laterally through the network until they find a dormant or orphaned account that allows them to escalate their privileges.

Unlike external hackers, insiders already start within the perimeter, while also benefitting from know-how of where sensitive assets and data lie and how to zero in on them. Insider threats take the longest to uncover—as employees, and other insiders, generally benefit from some level of trust by default, which may help them avoid detection. The protracted time-to-discovery also translates into higher potential for damage. Many of the most catastrophic breaches in recent years have been perpetrated by insiders.

Discover all the privileged accounts in your organization now with our free PowerBroker Privilege Discovery and Reporting Tool (DART). (CTA within glossary term)

Benefits of Privileged Access Management

The more privileges and access a user, account, or process amasses, the greater the potential for abuse, exploit, or error. Implementing privilege management not only minimizes the potential for a security breach occurring, it also helps limit the scope of a breach should one occur.

One differentiator between PAM and other types of security technologies is that PAM can dismantle multiple points of the cyberattack chain, providing protection against both external attack as well as attacks that make it within networks and systems.

PAM confers several chief benefits, including:

A condensed attack surface that protects against both internal and external threats:Limiting privileges for people, processes, and applications means the pathways and entrances for exploit are also diminished.

Reduced malware infection and propagation: Many varieties of malware (such as SQL injections, which rely on lack of least privilege) need elevated privileges to install or execute. Removing excessive privileges, such as through least privilege enforcement across the enterprise, can prevent malware from gaining a foothold, or reduce its spread if it does.

Enhanced operational performance: Restricting privileges to the minimal range of processes to perform an authorized activity reduces the chance of incompatibility issues between applications or systems, and helps reduce the risk of downtime.

Easier to achieve and prove compliance: By curbing the privileged activities that can possibly be performed, privileged access management helps create a less complex, and thus, a more audit-friendly, environment.

Additionally, many compliance regulations (including HIPAA, PCI DSS, FDDC, Government Connect, FISMA, and SOX) require that organizations apply least privilege access policies to ensure proper data stewardship and systems security. For instance, the US federal government’s FDCC mandate states that federal employees must log in to PCs with standard user privileges.

The more mature and holistic your privilege security policies and enforcement, the better you will be able to prevent and react to insider and external threats, while also meeting compliance mandates.

Here is an overview of the nine most important PAM best practices:

Establish and enforce a comprehensive privilege management policy:The policy should govern how privileged access and accounts are provisioned/de-provisioned; address the inventory and classification of privileged identities and accounts; and enforce best practices for security and management.

Identify and bring under management all privileged accounts and credentials:This should include all user and local accounts; application and service accounts database accounts; cloud and social media accounts; SSH keys; default and hard-coded passwords; and other privileged credentials – including those used by third parties/vendors. Discovery should also include platforms (e.g., Windows, Unix, Linux, Cloud, on-prem, etc.), directories, hardware devices, applications, services / daemons, firewalls, routers, etc. The privilege discovery process should illuminate where and how privileged passwords are being used, and help reveal security blind spots and malpractice, such as:

Orphaned accounts that could provide an attacker with a backdoor to your critical infrastructure

Passwords with no expiration date

Inappropriate use of privileged passwords—such as using the same Admin account across multiple service accounts

SSH keys reused across multiple servers

Enforce least privilege over end users, endpoints, accounts, applications, services, systems, etc.:A key piece of a successful least privilege implementation involves wholesale elimination of privileges everywhere they exist across your environment. Then, apply rules-based technology to elevate privileges as needed to perform specific actions, revoking privileges upon completion of the privileged activity.

Broken down to the tactical level, least privilege enforcement should encompass the following:

Remove admin rights on endpoints: Instead of provisioning default privileges, default all users to standard privileges while enabling elevated privileges for applications and to perform specific tasks. If access is not initially provided but required, the user can submit a help desk request for approval. Almost all (94%) Microsoft system vulnerabilities disclosed in 2016 could have been mitigated by removing administrator rights from end users. For most Windows and Mac users, there is no reason for them to have admin access on their local machine. Also, when it comes down to it, organizations need to be able to exert control over privileged access for any endpoint with an IP—traditional, mobile, network device, IoT, SCADA, etc.

Remove all root and admin access rights to servers and reduce every user to a standard user. This will dramatically reduce the attack surface and help safeguard your Tier-1 systems and other critical assets. Standard, “non-privileged” Unix and Linux accounts lack access to sudo, but still retain minimal default privileges, allowing for basic customizations and software installations. A common practice for standard accounts in Unix/Linux is to leverage the sudo command, which enables the user to temporarily elevate privileges to root-level, but without having direct access to the root account and password. However, while using sudo is better than providing direct root access, sudo poses many limitations with regards to auditability, ease of management, and scalability. Therefore, organizations are better served by employing server privilege management technologies that allow granular privilege elevation elevate on an as-needed basis, while providing clear auditing and monitoring capabilities.

Apply least privilege access rules through application control and other strategies and technologies to remove unnecessary privileges from applications, processes, IoT, tools (DevOps, etc.), and other assets. Enforce restrictions on software installation, usage, and OS configuration changes. Also limit the commands that can be typed on highly sensitive/critical systems.

Implement privilege bracketing – also called just-in-time privileges (JIT): Privileged access should always expire. Elevate privileges on an as-needed basis for specific applications and tasks only for the moment of time they are needed.

Limit privileged account membership to as few people as possible.

Minimize the number of rights for each privileged account

Enforce separation of privileges and separation of duties:Privilege separation measures include separating administrative account functions from standard account requirements, separating auditing/logging capabilities within the administrative accounts, and separating system functions (e.g., read, edit, write, execute, etc.). When least privilege and separation of privilege are in place, you can enforce separation of duties. Each privileged account should have privileges finely tuned to perform only a distinct set of tasks, with little overlap between various accounts. With these security controls enforced, although an IT worker may have access to a standard user account and several admin accounts, they should be restricted to using the standard account for all routine computing, and only have access to various admin accounts to accomplish authorized tasks that can only be performed with the elevated privileges of those accounts.

Segment systems and networksto broadly separate users and processes based on different levels of trust, needs, and privilege sets. Systems and networks requiring higher trust levels should implement more robust security controls. The more segmentation of networks and systems, the easier it is to contain any potential breach from spreading beyond its own segment.

Enforce password security best practices:

Centralize security and management of all credentials (e.g., privileged account passwords, SSH keys, application passwords, etc.) in a tamper-proof safe. Implement a workflow whereby privileged credentials can only be checked out until an authorized activity is completed, after which time the password is checked back in and privileged access is revoked.

Ensure robust passwords that can resist common attack types (e.g., brute force, dictionary-based, etc.) by enforcing strong password creation parameters, such as password complexity, uniqueness, etc.

Routinely rotate (change) passwords, decreasing the intervals of change in proportion to the password’s sensitivity. A top priority should be identifying and quickly changing any default credentials, as these present an out-sized risk. For the most sensitive privileged access and accounts, implement one-time passwords (OTPs), which immediately expire after a single use. While frequent password rotation helps prevent many types of password re-use attacks, OTP passwords can eliminate this threat.

Eliminate password sharing—each account should have a unique login to ensure a clear oversight and a clean audit trail.

Never reveal passwords—implement single sign-on (SSO) authentication to cloak passwords from both users and processes.

Remove embedded/hard-coded credentials and bring under centralized credential management. This typically requires a third-party solution for separating the password from the code and replacing it with an API that enables the credential to be retrieved from a centralized password safe.

Monitor and audit all privileged activity: This can be accomplished through user IDs as well as auditing and other tools. Implement privileged session management and monitoring (PSM) to detect suspicious activities and efficiently investigate risky privileged sessions in a timely manner. Privileged session management involves monitoring, recording, and controlling privileged sessions. Auditing activities should include capturing keystrokes and screens (allowing for live view and playback). PSM should cover the period of time during which elevated privileges/privileged access is granted to an account, service, or process. PSM capabilities are also essential for compliance. SOX, HIPAA, GLBA, PCI DSS, FDCC, FISMA, and other regulations increasingly require organizations to not only secure and protect data, but also be capable of proving the effectiveness of those measures.

Enforce vulnerability-based least-privilege access: Apply real-time vulnerability and threat data about a user or an asset to enable dynamic risk-based access decisions. For instance, this capability can allow you to automatically restrict privileges and prevent unsafe operations when a known threat or potential compromise exists for the user, asset, or system.

Implement privileged threat/user analytics: Establish baselines for privileged user activities and privileged access, and monitor and alert to any deviations that meet a defined risk threshold. Also incorporate other risk data for a more three-dimensional view of privilege risks. Accumulating as much data as possible is not necessarily the answer. What is most important is that you have the data you need in a form that allows you to make prompt, precise decisions to steer your organization to optimal cybersecurity outcomes.