Govern AI telemetry, systems, and evidencebefore risk spreads.
Cerbi AI Logging Governance gives Pro++ customers metadata-first visibility across connected AI logging paths: observed systems, providers and models, agent and tool context where emitted, request and token volume, estimated cost posture, and redacted evidence. It is separate from the Cerbi AI assistant.
Pro++ opt-in. Raw prompt and response storage off by default.
Start from one governed picture of connected AI logging paths.
See observed AI systems, providers and models, agent and tool context, and policy outcomes when those signals are present in scanner or governed runtime evidence.
Turn every governed event into a reviewable record.
Each governed AI logging signal can produce a bounded evidence packet: score, violation evidence, rule profile version, service and environment, and remediation context without storing the raw prompt or response by default.
A dashboard buyers can understand in minutes.
The flagship AI surface is split into focused views. Teams can start with source coverage, then move into runtime usage, payload exposure, and evidence without scrolling through one overloaded report.
CerbiShield / Sources
Show exactly how AI entered the governed path.
One view across connected AI logging paths.
Systems, providers, agents, tools, tokens, cost, exposure, policy, and evidence, governed when customers connect scanner or runtime telemetry paths to Cerbi.
Observed AI systems
Review named apps, agents, and tools from connected Cerbi evidence paths.
Providers & models
Track provider and model context by service and environment where metadata is present.
Agent and tool context
Surface tool and agent activity when applications emit supported telemetry fields.
Requests and tokens
Show request and token posture from metadata, not retained prompt text.
Estimated cost posture
Estimate usage impact from observed metadata; provider invoices remain authoritative.
Payload exposure
Flag raw prompt, response, RAG context, or tool payload logging risk.
Policy outcomes
Connect AI logging evidence to allow, redact, mask, hash, block, and audit decisions.
Evidence drilldowns
Move from score to rule match, digest, baseline movement, and remediation context.
Full visibility without the liability of storing raw prompts.
Every governed AI event produces a bounded evidence packet. It's what makes the governance story concrete: a reviewer gets what they need without CerbiShield becoming another place raw sensitive data lives.
Governance score
A single score summarizing policy conformance for the event.
Violation evidence
What triggered a finding, and the severity assigned to it.
Rule profile version
Which governance policy version evaluated the event.
Service & environment
Where the event originated, for routing and ownership.
Remediation context
What a reviewer needs to act on the finding.
Not the Cerbi AI assistant.
The Cerbi AI assistant helps teams operate Cerbi: rule writing, suggestions, explanations, and onboarding. AI Logging Governance is a different capability for connected AI telemetry metadata, scanner findings, runtime log observations, storage posture, and redacted evidence. The two are independent, and AI Logging Governance is disabled by default.
Metadata-first
Governance metadata and evidence, not raw content by default.
Raw off by default
Raw prompt and response storage requires explicit customer choice.
Customer-hosted
Deploys inside your own Azure tenant.
Observability stays in place
Doesn't replace your existing APM, SIEM, or logging tools.
What Cerbi does not do today.
These capabilities are on the roadmap, not implemented. Keeping this list separate from what's shipped is intentional: claims should follow evidence.
Universal enterprise-wide AI discovery
Automatic discovery across every cloud and internal system, not only integrated sources.
Autonomous agent action governance
Policy control over what autonomous agents are allowed to do, not only what they log.
Universal tool-call enforcement
Blocking or altering tool calls in-line, across any framework, not only observing them.
Model-behavior governance
Evaluating model outputs against behavioral policy, beyond telemetry and logging risk.
Cross-cloud execution governance
A single governance boundary spanning multiple clouds and runtimes.
Responsible AI / GRC lifecycle
Full risk, compliance, and governance lifecycle management for AI programs.
Pricing
AI Logging Governance is available in Pro++.
Pro++ is $499/month for up to 1B governed events, with AI Logging Governance available as an opt-in deployment capability. See the full plan model or talk to us about fit.
FAQ