Mate Security Unveils Gamebooks: Trusted Autonomy for AI-Powered Security Investigations
August 26, 2026 – 2:10 pm
Source: Mate Security
As artificial intelligence (AI) advances security operations toward greater speed and automation, the primary concern shifts from whether AI can investigate threats to whether organizations can trust it. SiliconANGLE previously reported on Mate Security’s Gamebooks, a novel architectural layer designed to provide AI agents with the freedom to reason and adapt while maintaining investigations within the organization’s established guidelines.
Moving Beyond Static Security Playbooks
Mate Security’s Gamebooks address a common challenge faced by security teams adopting AI: traditional SOAR (Security Orchestration, Automation, and Response) investigation playbooks are brittle and require continuous updates as threats, environments, tools, and business processes evolve. Conversely, AI SOC (Security Operations Center) platforms offer more flexible agentic reasoning but unbounded agents can be difficult to trust in real-world systems.
Mate’s solution? Separate investigative intent from specific execution steps. Gamebooks define:
- What must be investigated
- What evidence is needed
- Conditions that alter the investigation
- Permitted actions and escalation points
- When an agent should stop, escalate, or request approval
Unlike conventional playbooks, Gamebooks describe investigative intent rather than a fixed path, allowing agents to tailor their approach based on uncovered evidence and the organization’s current context.
A Layered Architecture for Agentic Investigations
Gamebooks build upon Mate’s existing Security Context Graph and Continuous Detection/Continuous Response (CD/CR) framework:
- Security Context Graph: Provides organizational context for agent reasoning.
- CD/CR Framework: Connects detection, investigation, and response into a continuous loop.
- Gamebooks: Establish investigative intent, required evidence, and boundaries.
- Capabilities: Reusable, vendor-neutral security skills dynamically applied as evidence emerges.
- Agents: Interact with tools and systems through the Flows layer, maintaining controlled execution.
The architecture’s goal is to enable execution to evolve without altering the underlying investigative methodology, ensuring adaptability in changing enterprise environments.