T Techclick ← All lessons
Zscaler · ZIA · Interactive lesson

How Zscaler controls GenAI SaaS

Your downloaded class poster, as a lesson: ChatGPT still hits ZIA. Inspect first. Then Cloud App Control (AI & ML), DLP, and Allow / Caution / Block / Isolate. Do not mix AI Guard, DSPM, or CASB at-rest into that inline path.

12 min read · L2 primary · Quiz at end

After this page you can

Lessons · Cloud App Control · GenAI SaaS

Infographic · Day 02 / ZIA+AI · 10 min
How Zscaler controls GenAI SaaS: user to connector to ZIA inspect then Allow Caution Block Isolate
There is no magic AI button. Traffic still hits ZIA. Name the objects: Cloud App Control, SSL Inspection, DLP, Isolate, AI Guard.
Say this out loud

Caution is the coach. Isolate is the remote browser. DLP is the paste-stop. AI Guard is runtime for YOUR models. Do not mix the boxes.

Why this exists

Users open ChatGPT in a browser tab. Prompts, file uploads, and pasted source code leave the company over HTTPS. ZIA can see that traffic only if it is forwarded and decrypted. If SSL Inspection is Do Not Inspect, Cloud App Control and DLP see a domain name, not the prompt.

The poster is the class drawing: one inline path, four actions, three inspection engines, four other official layers that are not that inline path.

The inline path

Read left to right:

  1. User / browser or ChatGPT tab — any device. Prompt and response are still HTTPS.
  2. Zscaler Client Connector or PAC — ZCC is the recommended forwarder. PAC also works. Traffic must reach ZIA.
  3. ZIA cloud / Zero Trust Exchange — all forwarded traffic terminates here. Policy enforcement point. Global anycast.
  4. SSL/TLS Inspection Policy = Inspect — decrypt so later engines can read the prompt. Required for inline controls.
  5. Cloud App Control → DLP — then an action: Allow, Caution, Block, or Isolate.
Fail path — limited visibility

SSL/TLS Inspection = Do Not Inspect → Cloud App Control / DLP cannot see the prompt → only domain-level Allow or Block is possible. That is why “we blocked chatgpt.com” still leaks if the user hits another AI app you never classified.

Inspection stack (three layers)

  1. 1 · Cloud App Control — AI & ML Applications

    Path: Policy → URL & Cloud App Control → Cloud App Control Policy → Add → AI & ML Applications.

    Use the predefined ChatGPT app (or the AI & ML category). Granular ChatGPT actions include Chatting, Uploading, Downloading, Deleting, Sharing, Inviting.

  2. 2 · Inline DLP

    Path: Policy → Data Loss Prevention → Rule With Content Inspection. This is the paste-stop for source, PII, PCI, PHI.

  3. 3 · DLP engines + dictionaries

    Source Code, PII/PCI/PHI class, Exact Data Match (EDM), Indexed Document Match (IDM) if you licensed them. Without a dictionary, “block secrets in ChatGPT” is a slogan.

Allow / Caution / Block / Isolate

ActionWhat the user feelsWhen to use
AllowFull access (optionally with granular ChatGPT actions).Approved users, after Inspect + DLP are on.
CautionEnd User Notification (EUN) coaches them, then access. Officially only for CONNECT, GET, or HEAD — one method per Caution rule.Teach first. Not a remote browser.
BlockDenied.Shadow AI you will not permit.
IsolateRemote browser (Zero Trust Browser / Cloud Browser Isolation). Download/paste stay off the endpoint.Need the app, not the data on disk. Requires an Isolation Profile. Isolate exists only if Isolation is enabled.

Other official layers — not the inline path

A · AI Guard

Runtime guardrails on prompts/responses for enterprise-managed AI / LLMs (prompt injection, data in prompts). Separate from public chatgpt.com Cloud App Control.

B · DSPM

Finds and classifies sensitive data sitting in cloud / SaaS / AI stores. Not the inline paste-block.

C · CASB / SaaS Security — Data at Rest

DLP for a ChatGPT Enterprise tenant: SaaS Application Tenants → ChatGPT → Data at Rest Scanning → Gen AI. Discovers data already in the tenant.

D · Analytics → Gen AI Security Reporting

Sanctioned vs unsanctioned (Shadow AI), prompt visibility if Gen AI Prompt Configuration is enabled.

Do this in the tenant

https://admin.zscaler.net
Training mock · not live

Policy → URL & Cloud App Control → Cloud App Control Policy → Add → AI & ML Applications

Add AI & ML rule

ChatGPT (AI & ML Applications)
Caution (EUN) — or Isolate if Isolation Profile exists
SSL/TLS Inspection Policy action = Inspect for this app

Source: Zscaler Help — Adding an AI & ML Applications Rule for Cloud App Control. Training mock, not a live tenant.

  1. Inspect first

    Policy → SSL/TLS Inspection → SSL/TLS Inspection Policy — Inspect for ChatGPT / AI & ML. Without this, stop.

  2. Write the AI & ML Cloud App Control rule

    Add rule, pick ChatGPT, set Allow / Caution / Block / Isolate. If Isolate, attach Isolation Profile.

  3. DLP on the same users

    Content inspection rule with Source Code + PII dictionaries. Test paste of a dummy secret.

  4. Prove it

    Web Insights Logs: user, cloud app = ChatGPT, SSL Inspected = Yes, policy action. Then Analytics → Gen AI Security Reporting if licensed.

Traps

SymptomLikely missFix
Logs show chatgpt.com but no promptDo Not InspectInspect that app, then retest.
Caution never fires on POST chatCaution is limited to CONNECT / GET / HEADUse Block, Isolate, or granular ChatGPT actions for chat/upload.
Isolate greyed outIsolation not enabled / no profileEnable Cloud Browser Isolation; attach Isolation Profile.
“AI Guard will stop ChatGPT.com”Wrong boxAI Guard is for your hosted models. Public ChatGPT is Cloud App Control.
Files already in ChatGPT EnterpriseInline DLP never saw themCASB Data at Rest Scanning on the tenant.

Knowledge check

Six judgement questions. If you mix AI Guard with Cloud App Control, redo the poster.

Q1

A Cloud App Control rule for ChatGPT exists, but Web Insights never shows the prompt. What failed first?

Correct: b. Inline Cloud App Control and DLP need decrypted traffic. Do Not Inspect leaves only domain-level Allow/Block.
Q2

Which sentence is true?

Correct: a. Do not mix the boxes. Caution coaches. Isolate relocates the browser. DLP needs inspect.
Q3

Where do you add the ChatGPT Cloud App Control rule?

Correct: a. Official Help path for an AI & ML Applications rule.
Q4

A user pastes source code into ChatGPT. Which control is the paste-stop on the wire?

Correct: b. Reporting sees yesterday. DSPM sees stores. Inline DLP is the paste-stop.
Q5

Isolate is selected but the rule will not save. What is missing?

Correct: b. Isolate exists only if Isolation is enabled. Attach the Isolation Profile.
Q6

ChatGPT Enterprise already holds uploaded files. Which layer looks at data at rest in that tenant?

Correct: c. Inline path never saw those files. Scan the tenant.

Sources

Related: URL Filtering + Cloud App Control · SSL Inspection · ZIA DLP · ZIA traffic flow