Sekoia Details How Its SOC AI Handles Customer Data
Summary
Sekoia says its SOC AI processes data within its service environment, does not use customer data to train generative models or send it to third-party AI providers, and operates under platform controls with analyst oversight.
Key points
- Sekoia says customer data is not used to train its generative foundation models and is not sent to third-party AI providers.
- The company says AI processing occurs within its applicable service environment; Elevate’s agentic models are hosted on Sekoia infrastructure.
- AI features query authorized platform data directly and, according to Sekoia, do not store customer data.
- Access and actions are governed by permissions, connectors, tenant settings, runbooks, and platform controls.
- Sekoia says customer environments are logically isolated, with encryption, action logging, consumption limits, and prompt-injection protections.
- Analysts can inspect investigation evidence and queries, then correct or replace proposed verdicts; configured high-impact actions require human validation.
- Data location, retention, and deletion depend on the service region, subscription, feature, and contract.
Article Details
- Topic
- Sekoia's approach to customer-data protection, transparency, and human oversight in AI-powered SOC operations.
People
Vendors
Products
AI CasesAI CasesElevateElevate queries, it doesn’t copy. It queries platform data directly and doesn’t store customer data.RoyRoySekoia autonomous SOCsit inside the architecture behind our agentic cybersecurity operations. AI works within the Sekoia autonomous SOC platform, with controlled access to customer data and a clear record of the investigation it conducts.