Protect my own device
For an individual trying it out. Open the app, sign in, turn it on, about
two minutes, no technical setup.
Roll it out to my company
For IT and security leaders. Push it to every employee’s device and watch
all AI activity in one dashboard.
Before you begin
- A PromptGuard account. You sign in to it from the app in step 3; if you don’t have one yet, create it on that sign-in page. No API key is needed.
- A supported computer. macOS 12 or later (Apple Silicon or Intel; the network extension, which also covers apps that ignore the system proxy, needs macOS 13), Windows on x64, or Linux on x64 or arm64. See the platform table for each installer.
- Permission to approve system prompts, once: a Mac asks you to allow the
network extension and to trust the inspection certificate with your
password, Windows shows a security prompt for the certificate, and the Linux
.deband.rpminstall withsudo. - Outbound HTTPS to
api.promptguard.co(scans and sign-in) andapp.promptguard.co(the browser sign-in page). Updates also fall back to PromptGuard’s GitHub releases (Updates).
Protect your device
About 5 minutes: download → install → sign in → turn on protection. Then try the prompts under See it work.1
Download the app
Grab the installer for your platform from
promptguard.co/download: no waitlist,
no sign-up required to download.
- macOS: a universal
.dmg(one installer for Apple Silicon and Intel), signed and notarized by Apple. - Windows: an
.exeinstaller (an.msiis also available for admins). The installers are code-signed by PG TECH LTD. If SmartScreen still warns on first run, it is two clicks, below. - Linux: an
.AppImage, a.debor an.rpmpackage (x64 on the download page; arm64 packages from support). Only the.deband.rpminstall the service that also captures AI tools which ignore the system proxy; see Supported AI tools. The.rpmisn’t on the download page yet: ask support@promptguard.co for it.
2
Install it
- macOS
- Windows
- Linux
Double-click the
.dmg and drag PromptGuard Shadow into
Applications (it has to be there for macOS to run its network
extension). When you first turn protection on, macOS asks you two
things, once each: to allow PromptGuard’s network extension in
System Settings (the app shows a button that opens the right
page), and to trust its HTTPS-inspection certificate with your Mac
password. The network extension needs macOS 13 or later; on macOS 12
the app uses the system proxy instead, which fewer apps follow.3
Sign in
Click the PromptGuard shield in your menu bar, then Connect this
device. Your browser opens, you sign in with your work account, and you’re
returned automatically, no keys or codes to copy.
4
Turn on protection
Click Turn on protection. You’ll be asked once to trust the secure
inspection certificate (and, on a Mac, to allow the network extension).
After that it runs quietly in the background, for as long as the app is
running: quitting the app stops the checking.
5
You're protected
The shield reads “You’re protected.” That’s it, keep working in ChatGPT,
Claude, and the rest, exactly as you do today.
See it work
In ChatGPT (or Claude), try these and watch what happens:
Click the shield → Activity to see every decision, why it was made, and a
one-click way to copy a safe version of anything that was blocked. The chips at
the top of Activity count each outcome: Allowed, Redacted and
Blocked.
The same events reach your organization’s dashboard: open
app.promptguard.co and go to Shadow AI, which
lists what was blocked or redacted before it left the device.
Deploy to your company
You don’t install Shadow AI machine-by-machine. An admin rolls it out once and manages the whole fleet from the dashboard:1
Create an enrollment token
In the dashboard, open Shadow AI and click Create a rollout token (MDM).
This one token enrolls many devices, up to its use count and lifetime.
2
Push it to your devices
Deploy the app through your existing MDM (Jamf, Intune, Kandji, …) with the
token in its managed configuration.
There is no sign-in prompt, and on a Mac the MDM profiles pre-approve the
network extension. One prompt remains: the first time protection turns
on, each person is asked to trust the device’s inspection certificate (their
Mac password, or a Windows security warning). Every device creates its own
certificate, so MDM can’t approve it ahead of time. Tell people to expect it.
3
Watch everything in one place
Every block, redaction, and allowed prompt, across browser and desktop,
rolls up to your dashboard, attributable per employee, revocable per device.
Full fleet rollout guide
Tokens, per-device credentials, attribution, and instant revocation.
Turn it off
Click the shield → Turn off protection (or, while paused, it resumes automatically). To remove it, see Uninstall: removal takes the certificate and local data with it, except that uninstalling the Windows.msi doesn’t yet, so run pgshadow uninstall first there.
Advanced: install from the command line
Advanced: install from the command line
The app carries a command-line tool, Managed fleets don’t need a command at all: put an enrollment token in the
pgshadow, for scripts and RMM tools. There
is no separate installer script: install the app as above, then call pgshadow
from inside it (it isn’t on your PATH on macOS or Windows):enrolment_token managed key and the device enrolls itself on first launch. See
the Desktop Agent reference for every
command.Next steps
Roll out to a fleet
Enrollment tokens, MDM configuration, per-device revocation
Evaluate Shadow AI
A proof of concept on one test device, with pass/fail criteria
Customise policies
Choose what is blocked, redacted or allowed
Supported AI tools
Which apps and sites are covered on each platform