What is Lumen?
Lumen is an AIDR — AI Detection & Response service. It gives an organization visibility into, and control over, how its people and its applications use generative AI.
A family of collectors (the sensor) captures AI interactions across browsers, internal applications, agentic and MCP traffic, API gateways, and cloud AI platforms. Each interaction is inspected in real time for LLM-specific risks, and Lumen can Log, Redact, or Block it per policy. Telemetry and findings live in a multi-tenant console and export to the Wazuh Indexer for correlation with the rest of your security telemetry.
Two deployment postures
- Workforce monitoring. The organization owns the endpoints (Intune, Jamf, Workspace ONE) and deploys the endpoint agent — with its capturing proxy and Claude Code hook guard — and the browser extension to managed machines. This is where the locally installed sensor lives, and it is the emphasis of the product.
- AI application development. The organization builds AI systems and instruments them inline with the Application SDK, the MCP proxy, a gateway plugin, cloud ingestion, or OpenTelemetry. No host agent required.
One tenant can run any mix of collectors. Every collector is registered in the console, associated with a policy, and given a tenant-scoped identity at enrollment.
What it detects
Six detection classes, mirroring the risks specific to LLM traffic:
| Class | Examples |
|---|---|
| Prompt injection & jailbreak | instruction override, system-prompt exfiltration, base64/homoglyph obfuscation |
| Sensitive-data exposure | credentials and secrets, PII, financial data, org-confidential identifiers |
| Malicious entities | phishing URLs, C2 domains, known-bad IPs in prompts or responses |
| Toxic / harmful content | violent, abusive, hateful, or self-harm content |
| Language | allowlist / denylist by detected language |
| Topic violations | configurable category restrictions (legal advice, medical advice, …) |
Details in Detection classes.
How it responds
Every interaction resolves to one effective action:
- Log — pass through and record.
- Redact — replace the matched spans before the model (prompt) or the user (response) sees them. The interaction proceeds.
- Block — stop the request or response and return a policy message.
The verdict is computed locally, inline, in single-digit milliseconds — the cloud never sits in the request path. See the policy model.
Part of the Wazuh Labs ecosystem
Lumen is a Labs-operated service: shared login with Wazuh ID, one organization across every Labs service, and billing through the shared customer record. Findings stay in Lumen's own data plane — the Hub only ever sees activation state and headline counts, never prompts, responses, or findings.