Plug and play DLP controls for your AI powered users.
Works with ChatGPT·Claude·Gemini
Live dashboard
The Intel dashboard deduplicates noisy alerts and scores each incident 0–100. Only genuine threats rise to the top.
Critical
3
High
8
Medium
14
Suppressed
47
AKIA4MFECR8OQZPK1234 found in prompt…
"4532 1234 5678 9010" detected in file upload…
"234-56-7890" matched in pasted text…
47 low-signal alerts suppressed · 68% noise filtered
Use cases
Intercept outgoing messages before they reach any AI model — block or alert on PII, credentials, and confidential content in real time.
Every DOCX, XLSX, PPTX, PDF and text file uploaded to an AI tool is extracted and scanned against your DLP rules before it leaves the browser.
AWS keys, API tokens, passwords, and secrets are caught by pattern matching before employees accidentally paste them into an AI prompt.
Social security numbers, credit card numbers, phone numbers, and email addresses are flagged automatically without any custom configuration.
Finance, Legal, and Engineering can each have stricter or looser rulesets. Policies are scoped per department and applied automatically.
Every message and file that triggered a rule is logged with the user, platform, timestamp, and matched rule — ready for your next audit.
How it works
Deploy the browser extension to your team. Supports Chrome and Edge. Each download is pre-configured for your organisation — no manual setup.
10 standard DLP rules ship automatically. Add custom regex or keyword patterns. Scope stricter rules to specific departments in minutes.
Every prompt and file across ChatGPT, Claude, and Gemini is scanned. Violations are blocked or logged instantly — no VPN, no proxy.
Under the hood
The Chrome extension patches window.fetch and XMLHttpRequest on AI tool pages. When a message is submitted, the extension intercepts it before the network call completes, scans against your DLP rules using pattern matching, and either cancels the request (block) or lets it through (alert + log). No traffic passes through Slopfence servers. Only alert metadata — never full message content — is sent to your dashboard.
Features
Patches fetch and XHR at the browser level — zero latency, zero proxy.
Extracts text from Word, Excel, PowerPoint and PDF files via JSZip + pdf-parse.
Per-rule severity — block silently cancels the request; alert logs and continues.
Detects Confidential, Restricted, Internal and other markers in file content.
Every org starts with SSN, card numbers, AWS keys, PII patterns and more out of the box.
Deduplicates noise and surfaces only genuine threats with a 0–100 risk score.
Pricing
No per-seat fees. Pay for your org, not per head. All plans include a 14-day free trial.
Starter
Up to 25 users
Growth
Up to 100 users
Enterprise
Unlimited users
FAQ
Make sure they're doing it safely.
Get started freeNo card required · 14-day free trial · Cancel anytime