Shadow AI Discovery: Which Apps and Which Users Are Using AI
AI Security Audit · Page 4 of 15

Shadow AI Discovery: Which Apps and Which Users Are Using AI

Six sources of evidence, one inventory. Cloud-app discovery, firewall and DNS, browser extensions, OAuth consents, endpoints and expense reports.

Shadow AI Discovery: Which Apps and Which Users Are Using AI
AI Security Audit · Page 4 of 15

43% of incidents involved shadow AI

IBM's 2026 report more than doubled the number from 2025, and 68% of breached organizations had no way to detect it.

Shadow AI Discovery: Which Apps and Which Users Are Using AI
AI Security Audit · Page 4 of 15

An inventory is the first control

You cannot protect, log or contract for a tool you do not know about. Discovery is where every other AI control starts.

25+ yearsIT & cybersecurity experience
Certified expertsCISSP · C|CISO · Microsoft · Cisco
Southern CaliforniaLocal and on-site, based in Irvine
NationwideRemote audits & advisory
Schedule a consultation

Where AI use shows up

Nobody fills in a form when they sign up for a free AI tool. The evidence is in traffic, consents and devices. A good discovery pulls six sources together, matches them to users and data types, and produces a list the business can act on: approve, tolerate with conditions, or block.

Where AI use shows up

Six places shadow AI shows up

Cloud app discoveryDefender for Cloud Apps: AI apps, users, uploads, risk score
Firewall and DNS logsAI domains by device, including unmanaged devices
Browser extensionsAI assistants reading every page
OAuth consentsAI apps with access to mail and files
Endpoint inventoryInstalled AI apps and local models
Expense recordsPaid subscriptions on cards
One AI inventoryApp · users · data types · source · risk score · decision · owner

Discovery sources and what each one tells you

SourceWhat it revealsTool we useBlind spot
Cloud app discoveryWhich generative-AI apps users reach, risk score, upload volumeMicrosoft Defender for Cloud Apps, firewall app controlPersonal devices off the network
Firewall and DNS logsAI domains by device and time, including unmanaged devices on Wi-FiMeraki, Fortinet, Palo Alto, DNS filteringEncrypted traffic without app identification
Browser extensionsAI assistants reading every page, including webmail and CRMIntune and Edge/Chrome policies, endpoint inventoryUnmanaged browsers
OAuth app consentsAI apps granted access to mail, files and calendars via "Sign in with Microsoft"Entra ID enterprise applications, consent logsConsents on personal accounts
Endpoint software inventoryInstalled AI desktop apps, local models, coding assistantsIntune, Defender for Endpoint, RMMPortable apps
Expense and procurement recordsPaid AI subscriptions bought on cardsFinance exportsFree tiers
Shadow AI discovery

From list to decision: approve, tolerate, block

ApproveBusiness plan, SSO, contract reviewed, in the catalog
Tolerate with conditionsInternal data only; review date set; owner named
BlockUnsanctioned in Defender for Cloud Apps and on the firewall; alternative offered

A page from the sample report

Sample report · Calder Ridge Partners (fictional company) · illustrative data
Domain
Shadow AI
Finding
31 generative-AI apps discovered, 4 approved
Severity
High
Evidence
Cloud app discovery over 30 days shows 31 generative-AI apps in use by 112 users. Four are approved. Eleven have a Defender risk score below 5. Two browser extensions with full page access are installed on 46 devices.
Fix
Publish the approved-AI catalog, mark the 11 low-score apps unsanctioned, remove the two extensions by policy, and route new requests through the approval form.

Every finding in the real report has this shape: domain, finding, severity, evidence, fix. See the full sample report.

Ali Hassani, CISO, OC Security Audit
CISO-led advisory

Guided by Ali Hassani, CISO

Certified CISO (C|CISO) and CISSP with 25+ years in IT, cybersecurity, compliance and infrastructure, and dozens of Microsoft and Cisco certifications. Every AI Security Audit is read-only, evidence-based, and written so an owner and an administrator can both act on it.

Questions business leaders ask

Can you find AI use on personal phones?

Only when they use company Wi-Fi or a company account. Discovery is honest about its blind spots; the policy and training pages cover the rest.

Will discovery slow down our network or users?

No. Every source is a log or an inventory you already collect. The audit reads them; it does not install agents or inspect content.

How often should we re-run discovery?

Monthly for the cloud-app report, quarterly for the full inventory. The vCISO service does this for you.

Implementation partner

OC Security Audit stays independent and read-only. If you want hands-on administration, IT Perfection can implement the fixes: DLP rules, Conditional Access, app blocking, agent identities, and ongoing Microsoft 365 support.

IT Perfection managed IT

Want a real list of the AI tools in your business?

The AI Security Audit runs all six discovery sources and hands you an inventory with owners, data types and a decision for each app.

Read-only audit. No changes to your systems. Irvine, Orange County and Southern California on-site; remote nationwide.