Turn unknown AI use into an inventory you control.
Oriant names the AI provider and employee from endpoint traffic, then lets you sanction or block each provider. It never stores prompts, files, or message bodies.
See real AI use, not installed software
Oriant names the provider and employee from real outbound traffic, including use outside the browser.
Connect a machine
shell
curl -fsSL https://app.oriant.app/connect
Activity is attributed to the identity stamped during installation. To set it explicitly, append --email you@company.com to the end of the command above.
Command generated
Single-use code ready
Run on your machine
Waiting for you to run the command...
You're live
Waiting for the first event to land...
Single-use - expires in ~15 min.
From blind spot to first evidence in 25 minutes
Start with one Mac. Roll out across the fleet through your existing MDM after you see the value.
5 minCreate the tenant and generate a connection command.
10 minInstall on one machine and stamp its identity.
25 minReview the first named providers and choose what is sanctioned, unsanctioned, or blocked.
Give security and GRC an AI audit trail
Export who used which AI provider, when, and what policy action applied without storing prompts, files, or message bodies.
Provider
Identity
Action
ChatGPT
a.toller@acme.co
Allowed
Claude
m.haan@acme.co
Allowed
GitHub Copilot
s.lund@acme.co
Allowed
Mistral
j.berg@acme.co
Blocked
Perplexity
r.holm@acme.co
Allowed
ChatGPT
e.vogt@acme.co
Allowed
Claude
n.aas@acme.co
Allowed
Mistral
t.roe@acme.co
Blocked
GitHub Copilot
l.munk@acme.co
Allowed
Perplexity
k.dahl@acme.co
Allowed
No AI detection rules to maintain
Oriant maintains the provider catalog. Your team only decides which providers are sanctioned or blocked.
No hand-built provider blocklist
No prompt-classification rules
No separate analytics pipeline
No daily tuning to keep discovery working
See AI use across browsers, IDEs, terminals, and native apps
Endpoint detection covers traffic that browser extensions and SaaS inventories miss.
Browser
IDE
Terminal
Shadow AI questions, answered
How Oriant discovers and controls unapproved AI tools without reading employee prompts.
Shadow AI is employee use of AI tools that security has not sanctioned. Oriant finds this use from real endpoint traffic and identifies both the AI provider and the employee.