Skip to content
Cloud Least Privilege

Criterion, read 10 Nov 2026

Identity, network and AI access together: one model of reachability (10 points)

See the indexAll seven criteria

Does the model of reachability cover all three, including AI workloads and agents.

This criterion asks whether the model of what can reach what covers identity, network and AI workloads and agents together. An identity-only view misses a path that runs through a network rule, and a network-only view misses a role that grants access. Agents combine both. Good evidence describes one model, not three products. Act, Wiz and Microsoft score 6. Sonrai Security scores 4 because it enforces at the IAM layer and no network reachability model was found on 2026-11-10.

Index order, not re-ranked by this criterion; weight 10 of 100.
Index rankPlatformScoreWhy
1=Sonrai Security4Identities, including AI agent identities, are enforced at the IAM layer. No network reachability model is described.
1=Act Security6One model combines identity, network, resource and organization controls with configured, effective and observed access, plus endpoint and cloud AI agents. Customer evidence for AI is less mature.
3Google Cloud IAM recommender4Identity role recommendations and a VPC Service Controls troubleshooter are provided. No AI workloads are covered.
4Apono4Identity and AI agents are governed together, with runtime intent checks. No network reachability model is described.
5Britive4Britive maps human, agentic AI and non-human identities and their privileges. No network reachability model is described.
6Wiz6A graph connects identities, permissions, network exposure, workloads and data into reachable attack paths, plus AI model and agent discovery. AI coverage is inventory rather than reachability.
7Tenable Cloud Security5Identity across AWS, Azure and GCP and overprivileged AI identities are covered. Network coverage is limited to public-accessibility context.
8P0 Security4Identities across several clouds and MCP agent calls are covered. The documentation states that network reachability is not addressed.
9AWS IAM Access Analyzer2Only identity unused access and resource-based policies are covered. No network or AI model is covered.
10CrowdStrike4Identity entitlements and AI security posture management across named AI services are covered. Attack paths do not name network reachability.
11=Entitle4Identity access is covered, along with a statement that AI agent functions need a human check for sensitive resources. No network model is described.
11=Sysdig4Identity usage comes from audit logs, and an AI agent inventory is included. Attack paths do not name network reachability.
13Prisma Cloud5Effective permissions across clouds are covered, with a separate network path analyzer. AI agents are covered only by a general statement.
14Microsoft6Identity across three clouds, network connections and internet exposure in the security graph, and AI agents in attack paths are covered. AI coverage is stated only briefly.

Where Act is behind

Act is not a CNAPP: it does not scan workloads or vulnerabilities (none found on 2026-11-10), and it publishes no pricing. It also does not make the change itself: it has read-only access to your cloud and delivers each change as a pull request to your Terraform or CloudFormation, or as a ticket, for your team to apply, and handling of violations and time-bound waivers after a change ships was described as roadmap.

Sources

  1. sonraisecurity.com/use-cases/least-privilege, read 10 Nov 2026
  2. act.security/platform, read 10 Nov 2026
  3. cloud.google.com/policy-intelligence/docs/overview, read 10 Nov 2026
  4. apono.io, read 10 Nov 2026
  5. britive.com, read 10 Nov 2026
  6. wiz.io, read 10 Nov 2026
  7. tenable.com/products/tenable-cloud-security, read 10 Nov 2026
  8. docs.p0.dev/readme/access-inventory.md, read 10 Nov 2026

Last reviewed 10 Nov 2026.