Cloud identity visibility tools show which human and non-human identities hold permissions in your cloud accounts, what those permissions allow, and which ones get used. The category spans cloud-native analyzers, cloud infrastructure entitlement management (CIEM), and identity security posture management (ISPM), and each sees a different slice of the same estate.
The slice most likely to go missing is the person. A finding can show that an AWS role is over-permissioned and still leave the team asking who relies on it. Discovery comes before governance in the cloud too. A least-privilege program can only right-size the roles it can attribute, because nobody signs off on cutting access they cannot trace to an owner.
That makes attribution the test to run first in any evaluation, ahead of how many permissions a tool can count.
What Is Cloud Identity Visibility?
Cloud identity visibility answers three questions about a cloud account: which identities can act there, what each can do, and who answers for it. Those identities include employees signing in through the IdP, contractors on guest accounts, workloads, CI pipelines, and AI agents calling cloud APIs.
The second question is harder than it looks. Effective permissions come from identity policies, resource policies, role trust relationships, group membership, and inheritance down an organization hierarchy. NIST's General Access Control Guidance for Cloud Systems (SP 800-210, 2020) treats IaaS, PaaS, and SaaS separately because each service model manages access to different components.
Cloud identity visibility tools answer the first two questions from the provider's own data: policies, role bindings, and activity logs. The third answer sits outside the cloud, in the IdP, the HR system, and the records of whoever deployed a workload or agent.
Identity Blind Spots Across Cloud Environments and Identity Systems
Each blind spot below sits on a seam between a cloud provider and an identity system. Together they form the cloud share of the identity attack surface: access an attacker can use without breaking anything, because the identity behind it is valid.
Federated role sessions. Workforce users often reach AWS through federated access paths. The IdP authenticates the person, and AWS STS issues temporary credentials for a role. CloudTrail then logs each action under an assumed-role session, labeled with a session name the IdP supplies, typically a user ID or email address. A chained role session can carry a different name, though. AWS's answer is source identity, a value that persists through role chaining and cannot change mid-session. It works only when the IdP sends it and each connected role's trust policy allows sts:SetSourceIdentity.
Cross-account and pipeline trust. A role trust policy can let principals from another account, or tokens from an external OIDC provider, assume a role. For GitHub Actions, AWS recommends evaluating the token.actions.githubusercontent.com:sub condition so only specific workflows can assume the role (GitHub Docs). A wildcard written at the organization level lets workflows from any repository in that organization in.
Cloud-native service accounts. Google Cloud's documentation describes a service account holding the basic Editor role, with more than 3,000 permissions. Over 90 days it used only a few. The binding records what the account can do and nothing about which team runs the workload behind it.
Guest role assignments. In Azure, a contractor invited as a guest in Microsoft Entra ID can hold a role assignment on a subscription. If the contract ends and nobody removes the guest, the assignment stays valid, and the subscription's role assignments still list a name the HR system no longer knows.
Key Capabilities to Look for in Cloud Identity Visibility Tools
What separates cloud identity visibility platforms from point tools is whether they carry the analysis from inventory to owner to fix, across more than one provider. Test these capabilities in a proof of concept:
- Multicloud entitlement inventory. Every identity and grant across the cloud accounts you connect, including resource policies as well as identity policies.
- Effective-permission calculation. What an identity can do after inheritance, group membership, permissions boundaries, and organization policies apply. A list of attached policies is a starting point, and the math on top of it is the product.
- Usage-based right-sizing. Granted permissions compared with activity over a stated window, with the window shown. A job that runs once a quarter looks idle for most of the year.
- Federation and cross-account trust mapping. Which IdPs, external OIDC providers, and other accounts can assume which roles, and under what conditions.
- Non-human and agent identity coverage. Service accounts, workload identities, CI tokens, and AI agents inventoried as identities with their own entitlements and owners.
- Owner attribution. A named person or team for each identity, drawn from systems outside the cloud provider.
- Remediation workflow and evidence export. Findings routed to the owner, standing grants swapped for just-in-time access where workloads allow it, and an audit export of each change.
Cloud identity visibility tools draw the first three capabilities from data the provider already exposes, which makes them the easiest to demonstrate. Owner attribution depends on data the provider does not hold. Give it the longest test, in your own accounts.
Understanding CIEM and ISPM: Capabilities, Differences, and Overlap
CIEM works across cloud environments. It inventories identities and entitlements across IaaS and PaaS accounts, calculates effective permissions, compares them with usage, and recommends smaller grants. CIEM often ships as a module of a cloud security posture management (CSPM) product or a cloud-native application protection platform (CNAPP), next to misconfiguration scanning.
ISPM works on the identity systems themselves. Identity security posture management checks identity infrastructure for weak posture, starting with the IdP and directories and often extending to SaaS and cloud IAM. Typical findings are accounts without MFA, stale accounts, excessive admin roles, and permissive federation trust. Where CIEM asks what a cloud role can do, ISPM asks whether the identity configuration behind it is sound.
The overlap is privilege. Both flag excessive and unused privilege, and both touch the standing admin access that privileged access management programs also track. ISPM overlaps with identity threat detection and response (ITDR) too. ISPM primarily addresses standing posture weaknesses, while ITDR solutions detect and respond to active identity abuse, and many findings feed both.
| Category | Primary scope | Best for | Evaluate |
|---|---|---|---|
| Cloud-native tools | One provider's IAM | Unused roles, keys, and permissions in that cloud | Whether findings reach the identity's owner |
| CIEM | Entitlements across multiple clouds | Effective permissions, right-sizing, cross-account trust | Whether a federated session or guest grant resolves to a person |
| ISPM | IdPs, directories, SaaS, often cloud IAM | MFA gaps, stale accounts, admin sprawl | How far its view reaches into cloud roles and resource policies |
Pick the category by the question your team cannot answer. For "what can this cloud role do, and does it need it?", start with the native tools. Add a CIEM once several providers or deeper effective-permission analysis come into play. "Is the identity system behind this access sound?" is an ISPM question.
Connecting Identity Data Across Cloud and Identity Systems
Cloud identity visibility depends on two joins. The first runs from a cloud entitlement to an identity, and cloud identity visibility tools handle most of it. The second runs from that identity to an accountable owner, and it needs data the cloud provider does not hold.
The native tools show where the first join stops. AWS IAM Access Analyzer reports unused roles, unused IAM user access keys and passwords, and unused permissions within a usage window, billed per IAM role and user analyzed. An unused role's trust policy may name a SAML provider, though, and a SAML provider is not a person. Source identity closes part of that gap, but CloudTrail does not capture it when an AWS service or service-linked role acts on behalf of a federated identity.
Google's IAM recommender compares each principal's granted permissions with what it used in the last 90 days. Its lateral movement insights flag roles that let service accounts in one project impersonate service accounts in another, a lateral movement path. Google also notes that role recommendations consider only IAM access controls, leaving out access control lists and Kubernetes role-based access control.
The second join comes from the IdP, the HR system, and deployment records. For service accounts, CI tokens, and agents, that usually means an owner recorded when the identity is created, which is where non-human identity security programs start.
Whether to build the second join or buy it depends on how many systems hold part of the answer. Build it when one cloud provider and one IdP cover most access, and source identity plus consistent owner tags can carry attribution. Buy it when several IdPs, guest tenants, and agent platforms each hold a piece, since the reconciliation work grows with every system you add.
Choosing the Right Cloud Identity Visibility Solution
Choosing among cloud identity visibility solutions starts with the question your team cannot answer today. If unused access is the gap, turn on the native analyzers before buying anything, and use the findings they produce as the baseline every product must beat.
Weigh roadmap risk as well. Microsoft stopped selling Entra Permissions Management to new Enterprise Agreement and direct customers on April 1, 2025, and to new CSP customers on May 1. It retired the product on November 1, 2025, a month after the original date, and kept some CIEM features in Microsoft Defender for Cloud. A standalone product can be folded into a broader platform or retired, so ask how each vendor packages and licenses its entitlement features.
Shortlist a product only if it can resolve a federated session, a guest grant, and a service account in your own accounts to named owners. Then bring these questions to the demo:
- Which providers, account types, and resource policies does it read, and which does it leave out?
- How does it calculate effective permissions, and can it show the policy chain behind one result?
- What usage window does it apply, and how does it treat jobs that run once a quarter?
- How does it attribute a federated role session to a person when source identity is not set?
- Where does its owner data come from, and what does it show for an identity with no owner?
- Can it route findings to that owner and export the result as audit evidence?
The answers also show where cloud identity visibility tools end and a wider purchase begins. Teams weighing posture, detection, and governance together can set them beside an identity security platform comparison.
How NewCore Helps Map Identities Across Cloud and Identity Systems
The tools in the previous section can tell a team that a role is over-permissioned. They often cannot say whose it is, and without that answer nobody can approve the fix. NewCore's view is that many cloud stacks already handle the first join through native analyzers or a CIEM. The gap is the second one, from an entitlement to the person or agent accountable for it.
- One identity record: a person's accounts across identity providers, directories, and HR systems resolved into one record, with their AI agents linked to it. Risk reduced: cloud findings that cannot be routed to a person.
- Identity graph: people, AI agents, and applications in one connected view, with each access path traced across the systems you connect, including access outside SSO. Risk reduced: cloud access granted outside the IdP, which no IdP report shows.
- Agent directory: each agent found where it runs, with its owner, who can invoke it, and which tools and applications it can reach. Risk reduced: agents that reach cloud resources with no accountable owner.
- Ask NewCore: plain-language questions such as "Which agents can reach production, and who owns them?", with agents that have no owner flagged for review. Risk reduced: ownership answered by spreadsheet during an audit.
NewCore does not right-size cloud permissions or scan for misconfiguration, so it complements the cloud-native tools, the CIEM, and the IdP rather than replacing them. It adds people and AI agent discovery across connected systems, connecting those findings to the people and agents accountable for the access.
Every cloud role answers to someone. Know who before the next review. Request a Demo →
Frequently Asked Questions
Is CIEM part of CSPM or CNAPP?
Often, but not always. Many vendors sell CIEM as a module of a CSPM or CNAPP product, and standalone CIEM products also exist. Microsoft retired its standalone Entra Permissions Management in 2025 and kept some of its CIEM features in the Defender CSPM plan of Microsoft Defender for Cloud. Check which license tier includes entitlement analysis before assuming you already own it.
Do I need a CIEM if I already use AWS IAM Access Analyzer or the Google Cloud IAM recommender?
Possibly not, if you run one provider and the native findings reach the right owners. Native tools report unused roles, keys, and permissions within their own cloud. A CIEM can add a cross-provider view and broader effective-permission analysis. Turn the native tools on first, and treat the gaps they leave in your environment as your CIEM requirements.
What is the difference between ISPM and ITDR?
ISPM finds standing weaknesses in identity systems, such as missing MFA, stale accounts, or excessive admin roles, before anyone exploits them. ITDR detects and responds to identity abuse in progress, such as a suspicious session or an unexpected privilege change. They overlap: a posture finding often explains why a detection fired, and some products cover both.
Can cloud identity visibility tools see AI agents?
Partly. An agent that calls cloud APIs uses a role, service account, or token, so cloud tools see that credential and its permissions. They may not see the agent behind it, the person who owns it, or the other systems it can reach. Ask the vendor to trace one of your agents end to end during the evaluation.
Who should own cloud identity visibility: cloud security, IAM, or the platform team?
It depends on where decisions get made. Cloud security often runs the tooling, IAM owns identities and the IdP, and platform teams own the workloads. One workable split gives cloud security the program and makes IAM responsible for owner data. Whatever the split, name one team accountable for findings nobody can attribute, or those findings tend to stay open.
Discover every agent acting in your business.
Book a DemoIdentity Security, NewCore
The team behind NewCore's Identity 101 guides, covering agent security, identity security, workforce security, identity governance, and identity discovery. These guides draw on NewCore's own product and security work, translated into practical, vendor-neutral guidance for security and IT teams managing identity across people, software, and AI agents.
View all articlesCite this guide
The NewCore Team. (2026, November 18). Cloud identity visibility tools: a buyer’s guide to CIEM, ISPM, and identity security. NewCore. https://newcore.com/identity-101/cloud-identity-visibility-tools


