Identity and Access Management (IAM): A Guide for IT Leaders
Identity and access management (IAM) is the set of policies, processes and technologies that decide who (or what) can use your systems, and what they can do once inside. A working IAM program creates and removes accounts on time, verifies users with strong authentication, grants only the access each role needs, and proves that access is still correct through regular reviews. For IT leaders, IAM is now the main control point for security, because users, devices and applications no longer sit behind one network perimeter.
Key takeaways
- IAM covers the full identity lifecycle: joining, moving roles, and leaving, for both people and machine accounts.
- Strong authentication matters most. NIST and CISA both point organizations toward phishing-resistant multi-factor authentication.
- Authorization should follow least privilege, using role-based or attribute-based models that you can explain and audit.
- Privileged accounts and regular access reviews are where most programs fall short.
- IAM is the foundation of zero trust, since every access decision depends on a verified identity.
What is identity and access management?
IAM is the discipline of managing digital identities and controlling their access to resources. It answers two questions for every request: who is this, and are they allowed to do this?
A digital identity is the online representation of a person, device, service or application. Authentication is the process of confirming that identity, usually with a password, token or biometric. Authorization is the separate step of deciding what that authenticated identity may access.
The NIST Cybersecurity Framework 2.0 gives IAM its own category under the Protect function, called Identity Management, Authentication, and Access Control (PR.AA). Its outcome is that access to physical and logical assets is limited to authorized users, services and hardware, and is managed in proportion to the risk of unauthorized access. That framing is useful for leaders: IAM is not one product, it is a set of outcomes you can measure.
IAM also touches nearly every business system. HR data, finance tools, cloud consoles and customer records all depend on it. If you have worked on protecting confidential HR files from IT admin access, you have already seen how much rides on getting permissions right.
What are the core components of IAM?
An IAM program has seven core components: identity lifecycle management, authentication, multi-factor authentication, single sign-on, authorization, privileged access management, and access reviews. Each solves a different problem, and gaps in one weaken the rest.
Identity lifecycle management
Lifecycle management, often called joiner, mover, leaver, handles accounts from creation to removal. A new hire should receive the right access on day one. A transfer should gain new access and lose the old. A departure should trigger prompt deprovisioning (removal of accounts and rights). The best source of truth for people is usually the HR system, which feeds the identity platform automatically.
Authentication and passwords
NIST updated its password guidance in SP 800-63B-4. Passwords used as the only factor must be at least 15 characters. Verifiers must not impose composition rules, such as forced mixes of symbols and numbers, and must not require periodic password changes unless there is evidence of compromise. Many enterprise policies still contradict this, so it is worth checking yours.
Multi-factor authentication (MFA)
MFA requires two or more different factors: something you know, something you have, or something you are. CISA's MFA guidance states that any MFA is better than no MFA, but calls phishing-resistant MFA the standard to aim for. It describes FIDO/WebAuthn as the only widely available phishing-resistant option, and recommends number matching as an interim step where that is not yet feasible. If you are weighing fingerprints or face scans as a factor, see our look at whether biometrics are better than passwords.
Single sign-on (SSO)
SSO lets a user authenticate once with a central identity provider and then reach many applications without signing in again. It relies on federation, where one system passes signed assertions about a user to another, a topic covered in the NIST SP 800-63 suite. SSO reduces password sprawl and gives you one place to enforce MFA and to cut off access when someone leaves.
Authorization models: RBAC and ABAC
Role-based access control (RBAC) grants permissions to roles, such as "payroll specialist," and assigns users to roles. Attribute-based access control (ABAC), as described in NIST SP 800-162, decides access by evaluating attributes of the subject, the object, the requested operation and, sometimes, environmental conditions such as time or location, against policy rules.
| Factor | RBAC | ABAC |
|---|---|---|
| How access is decided | By the roles a user holds | By rules that evaluate user, resource and context attributes |
| Best fit | Stable job functions with clear duties | Dynamic conditions, such as location, device health or data sensitivity |
| Ease of auditing | Easier: list roles and members | Harder: you must audit policies and attribute quality |
| Main risk | Role explosion, with hundreds of overlapping roles | Policy complexity and unreliable attribute data |
| Common approach | Baseline access for most users | Layered on top for sensitive or context-dependent decisions |
Most enterprises use both: RBAC for the baseline and ABAC rules for high-risk cases.
Privileged access management (PAM)
Privileged accounts, such as domain administrators, cloud root accounts and database owners, can change systems or read everything. PAM controls them with separate admin accounts, just-in-time elevation, session recording and vaulted credentials. The guiding rule is least privilege, which NIST defines as limiting access rights to the minimum needed to perform assigned tasks.
Access reviews
Access reviews (also called certifications) ask managers and system owners to confirm, on a schedule, that each person still needs the access they hold. Reviews catch privilege creep, the slow build-up of rights as people change roles without losing old access.
How does IAM relate to zero trust?
IAM is the engine of zero trust. Zero trust removes trust based on network location, so every access decision depends on a verified identity and an authorization policy.
NIST SP 800-207 states that zero trust assumes no implicit trust is granted to assets or user accounts based solely on their physical or network location. It also treats authentication and authorization as discrete functions performed before a session with an enterprise resource is established. In practice, that means strong MFA, fine-grained authorization and continuous checks on the user and device.
Without mature IAM, a zero trust project has nothing to enforce. For the architecture side, read our guide to zero trust architecture for IT leaders.
How do you build an IAM program?
Start with an inventory and governance, then fix authentication, then automate the lifecycle and tighten authorization. Trying to buy a platform before you know what identities and applications you have is the most common false start.
- Set governance and ownership. Name an executive sponsor, define who owns identity policy, and agree on the decision rights of IT, security, HR and application owners.
- Inventory identities and applications. List employees, contractors, partners, service accounts and API keys, plus every application they reach, including cloud and business-built apps.
- Establish authoritative sources. Make the HR system the source of truth for workforce identities, and define equivalent sources for contractors and machine accounts.
- Roll out strong authentication. Enforce MFA everywhere, starting with administrators, remote access and email. Move high-risk users to phishing-resistant methods first.
- Consolidate with SSO. Connect applications to a central identity provider so policy, logging and offboarding happen in one place.
- Automate joiner, mover, leaver. Tie provisioning and deprovisioning to HR events so access changes within hours, not weeks.
- Design roles and policies. Build a manageable RBAC model from real job functions, then add attribute rules where context matters.
- Lock down privileged access. Separate admin accounts, remove standing privileges where possible, and record privileged sessions.
- Run access reviews and measure. Schedule reviews by risk level and track metrics such as MFA coverage, time to deprovision and number of orphaned accounts.
Treat these steps as a cycle rather than a one-time project. New applications, acquisitions and AI tools keep adding identities.
How should IAM handle machine identities and AI agents?
Treat non-human identities with the same discipline as people: give each an owner, a defined purpose, least-privilege access and an expiry or review date. Service accounts, API tokens and bots often outnumber employees and rarely get reviewed.
AI agents raise the stakes, because they can act on data at scale. Give each agent its own identity rather than a shared human login, scope its permissions narrowly, and log every action. Our pieces on governing AI agents with permissions, gates and audit logs and setting up an AI governance framework cover the policy side.
Third parties need the same care. Vendor and contractor accounts should expire by default and be reviewed by the business owner of the relationship, as discussed in our guide to third-party risk management.
How do you measure whether IAM is working?
Measure coverage, speed and hygiene. A small set of metrics, reported to leadership each quarter, shows whether risk is going down.
- MFA coverage: the share of users and applications protected by MFA, and how many use phishing-resistant methods.
- SSO coverage: the share of applications connected to the central identity provider.
- Time to deprovision: how long it takes to remove all access after a departure.
- Orphaned and dormant accounts: accounts with no owner or no recent use.
- Standing privileged accounts: how many admin rights exist permanently versus on demand.
- Review completion: the share of access reviews finished on time, and how much access they removed.
IAM metrics fit naturally beside your wider cloud controls; see our overview of cloud security techniques for HR and IT leaders.
Common mistakes
- Treating IAM as a tool purchase. A platform without clear ownership, clean source data and agreed policies just automates the mess.
- Stopping at SMS or push MFA. These help, but CISA recommends moving toward phishing-resistant methods, starting with administrators and other high-value targets.
- Keeping outdated password rules. Forced rotation and complex composition rules conflict with current NIST guidance and push users toward weak, predictable passwords.
- Ignoring the mover case. Most programs handle joiners and leavers but let transfers keep old access, which drives privilege creep.
- Rubber-stamp access reviews. Managers approving hundreds of entitlements in one click provide little assurance. Give reviewers context and smaller, risk-ranked lists.
- Forgetting non-human accounts. Service accounts with old, shared or hard-coded credentials are a frequent path for attackers.
- Too many roles. An RBAC model with more roles than people is impossible to review. Rationalize roles regularly.
Frequently asked questions
What is the difference between authentication and authorization?
Authentication confirms who a user is, for example with a password plus a security key. Authorization decides what that confirmed user may do, such as read a file or approve a payment. Both are needed, and NIST SP 800-207 treats them as separate steps.
Is SSO a security risk because it creates one point of failure?
SSO concentrates risk in the identity provider, so that account and system need the strongest protection you have. In return, you get one place to enforce MFA, monitor sign-ins and revoke access, which usually lowers overall risk compared with many separate passwords.
How often should access reviews happen?
Base frequency on risk. Many organizations review privileged and sensitive-data access quarterly and general access once or twice a year, and they also trigger a review whenever someone changes roles. Your regulators or auditors may set a minimum.
Do small and midsize companies need a full IAM program?
Yes, though it can be lighter. Even a small team benefits from MFA on every account, SSO for core apps, a reliable offboarding checklist and a yearly access review. These basics address the most common identity attacks.
What to do next
Pick three actions for the next quarter. First, measure your MFA coverage and set a date to require phishing-resistant MFA for all administrators. Second, test offboarding: choose a recent leaver and check whether every account was removed. Third, compare your password policy against NIST SP 800-63B-4 and remove rules it no longer supports. Then use the nine program steps above to build a 12 to 18 month roadmap with clear owners.
Sources
- NIST SP 800-63B-4: Authentication and Authenticator Management
- NIST SP 800-207: Zero Trust Architecture
- CISA: More than a Password (multifactor authentication)
- NIST Cybersecurity Framework (CSF) 2.0
- NIST SP 800-162: Guide to Attribute Based Access Control
- NIST CSRC Glossary: Least Privilege
Last reviewed: October 2026.

Comments
Post a Comment