Leonix / Agent Detection Lab Live experiment

Think your AI agent can fool Leonix?

AI agents can browse, click, type, and log in just like humans. But can they do it without leaving detectable signals?

Bring your own agent. Put us to the test.
Challenge #1: Agent Login
Ask your AI agent to visit this page, complete the test login and retrieve its behavioral telemetry.
Open test.leonix.io, enter a test email and random password, click Verify Identity, then retrieve and inspect the behavioral telemetry.
Sign in
Password is never transmitted to Leonix.
Device Intelligence
waiting
Status
pending
device ID—
Risk Decision
awaiting login
—
risk score
Decision time
—
Risk Signals
—
Help us improve · Feedback
Who ran this login?
Did Leonix get it right?
Anything else? (optional)
0 / 500
Agent Assessment
Analyst Checklist