Endpoint
Chat, coding agents, local tools
Where the sensor looks
AI is in the tools people already use.
Cortex finds sanctioned and shadow AI on supported capture paths: device, browser, IDE, SaaS, and MCP. Coverage is named, not assumed.
SaaS
Support, knowledge, and ops work
Code
PRs, agents, and review loops
Cloud
Hosted models and MCP
One event
One incident through Discover, Control, Investigate.
A sanctioned MCP read is denied on a Cortex-controlled path. The same payload is pasted into an unsanctioned ChatGPT tab and is observed only.
Discover · INC-SYNTH-01
Inventory and exposure
- Requester
- ACME-SUPPORT-01
- Surface
- Chrome · unsanctioned tab
- Model
- ChatGPT
- Data class
- pii.ssn
Control · INC-SYNTH-01
Decision on a managed hop
- Requester
- ACME-SUPPORT-01
- Surface
- Cursor MCP · approved client book
- Data class
- pii.ssn
- Destination
- MCP client-book read
Investigate · INC-SYNTH-01
Record the officer can defend
- Incident
- INC-SYNTH-01
- Requester
- ACME-SUPPORT-01 · Support · triage-helper
- Surface
- Chrome · unsanctioned tab
- Data class
- pii.ssn · SSN •••-••-0000
Deployment, lineage & assurance
One security program. Every session accounted for.
Run one evidence program across endpoints, cloud workloads, and self-hosted environments, with the deployment record and assurance posture in the same view.
- Which person or agent initiated the session
- Which surface, destination, and data class were involved
- Whether Cortex discovered, controlled, or investigated the event
Deploy where the session runs
Human-operated AI
Browser and desktop tools people already use. ChatGPT, Claude, Grok, Cursor, and the rest of the consumer surface.
On-device sensor
Cloud workloads
Hosted agents, provider APIs, and MCP paths: Slack agents, OpenAI, Gemini, GitHub.
Your cloud
Self-hosted
Run capture in your VPC or on-prem. Same inventory, same data classes, inside your boundary.
Your VPC or on-prem
Assurance carried with the evidence
SOC 2 Type IICompliantChaos Labs is SOC 2 Type II compliant. Monitored by Vanta.
ISO 27001CompliantISO/IEC 27001 information security management.
NIST AI RMFAlignedCortex is aligned to the NIST AI Risk Management Framework. Map and Measure start with which AI systems are in use and what data left the organization.
ISO/IEC 42001AlignedCortex is aligned to ISO/IEC 42001. A review of the AI management system asks for inventory of systems, data used, and third-party models.
Review your AI exposure.
Discover. Control. Investigate.