The Rise of AI Identities in SaaS Security
- Martin Snyder
- 14 hours ago
- 4 min read
AI systems are no longer just features.
They operate as identities inside SaaS environments.
If you’re not tracking AI identities, you’re not securing your perimeter.
AI Is No Longer Just a Feature
For years, security teams treated AI as functionality.
A productivity boost.
An analytics engine.
A summarization tool.
But in modern SaaS environments, AI is no longer just a feature.
It acts.
It connects.
It accesses.
It authenticates.
AI systems increasingly operate like identities.
And identities define your attack surface.
What Is an AI Identity?
An AI identity is not a human user.
It is:
An AI copilot with delegated file access
A summarization engine reading inbox content
A CRM assistant analyzing customer data
A workflow automation bot acting on behalf of users
A model-driven system with API keys
These AI-powered systems often authenticate via:
OAuth tokens
Service accounts
Application permissions
API keys
They have access.
They perform actions.
They process data.
From a security perspective, that makes them identities.
The Scale of the Problem
According to Waldo Security’s 2025 SaaS & Cloud Discovery Report:
97% of SaaS applications are unknown to IT
100% of organizations have unauthorized cloud accounts
Less than 1% of SaaS accounts enforce MFA
Now combine that with a simple reality:
Almost every modern SaaS platform now leverages AI.
If most SaaS apps are unknown, then most AI-enabled systems operating inside your organization are also unknown.
AI identities are expanding inside environments that were never fully mapped to begin with.
AI Identities Bypass Traditional Controls
AI systems rarely log in like employees.
They:
Use long-lived OAuth tokens
Operate through application permissions
Authenticate via service accounts
Sync data across platforms automatically
CISA’s Secure Cloud Business Applications (SCuBA) guidance highlights delegated access as a persistent risk because it can outlive users and evade centralized visibility:https://www.cisa.gov/secure-cloud-business-applications-scuba
When AI tools authenticate via delegated permissions, they inherit access silently.
They do not raise suspicion.
They look legitimate.
Because technically, they are.
AI Identities Change the Perimeter
The CISA Zero Trust Maturity Model centers identity as the foundation of modern security:https://www.cisa.gov/zero-trust-maturity-model
But traditional identity models assume:
Humans
Devices
Managed service accounts
AI identities blur those lines.
They:
Operate autonomously
Access data continuously
Analyze content in real time
Trigger actions without human review
If identity is the perimeter, AI has now expanded that perimeter.
Why AI Identity Risk Is Hard to See
AI systems often appear as:
“Connected apps”
“Integrations”
“Automation workflows”
“Enterprise applications”
They are rarely labeled as AI.
Security teams may see a connected app but not realize:
It processes files with machine learning
It analyzes sensitive communications
It retains data to improve model outputs
It connects across multiple SaaS platforms
If you are concerned about AI in your organization, discovery must start with SaaS.
Because AI lives inside SaaS.
And SaaS spreads at the speed of authentication.
Compliance Does Not Differentiate
Frameworks such as the NIST Privacy Framework and ISO/IEC 27001 require accountability for data processing activities:
If an AI system processes regulated or sensitive data, it must be governed.
It does not matter whether the processor is human or algorithmic.
Unknown AI identities represent undocumented data processing.
Undocumented processing represents exposure.
The Questions You Should Be Asking
For every AI-enabled SaaS platform, ask:
What identity does it use to authenticate?
What scopes or permissions are granted?
Does it have persistent OAuth access?
Does it process customer or regulated data?
Can access be revoked centrally?
Is there a business owner?
If you cannot answer those questions, the AI identity is unmanaged.
Unmanaged identities define modern breach paths.
From Human-Centric to Identity-Centric Security
Security programs historically focused on:
Employee lifecycle
Password policies
MFA coverage
Endpoint protection
Modern SaaS environments require a broader model:
Human identities
Non-human identities
Service accounts
OAuth tokens
AI-driven application permissions
All must be discovered.
All must be governed.
How Waldo Security Surfaces AI Identities
Waldo Security’s SaaS & Cloud Discovery Engine enables organizations to:
Discover known and unknown SaaS platforms
Surface OAuth and delegated access
Identify non-human and AI-driven identities
Detect Shadow cloud environments
Map SaaS and AI exposure to compliance frameworks
Because nearly every SaaS platform now leverages AI, understanding SaaS usage is inseparable from understanding AI identity risk.
You cannot govern AI if you cannot see the identities behind it.
Conclusion: AI Has Joined Your Identity Perimeter
AI is no longer just software.
It authenticates.
It accesses.
It analyzes.
It acts.
That makes it part of your identity perimeter.
If you are concerned about AI exposure, start with SaaS discovery.
Learn how organizations are uncovering SaaS and AI-related exposure in the 2025 SaaS & Cloud Discovery Report:
About Waldo Security
Waldo Security helps organizations discover, classify, and secure every SaaS and cloud service in use — known or unknown. By illuminating unmanaged identities, OAuth risk, Shadow IT, and AI-enabled SaaS exposure, Waldo enables security teams to defend the identity perimeter with continuous visibility and evidence.
Comments