Trust-based Adaptive Access Control Architecture
summary
Proposes an architecture that evaluates third‑party smart agents by running the agent in a sandbox alongside a digital twin simulation and doing live compliance checking, combining static analysis, digital signatures and a simulator to produce a computed trust score (example: 45% as a decision threshold). A trust aggregator feeds a decision tree that selectively exposes platform features to the agent. Prototype behavior is shown conceptually with sandboxed execution and reputation gossip inputs; not yet validated at cloud scale, suggested for Kubernetes/edge use.
AI-written summary · about
chapters (9)
- 0:00Trust-based access control intro
- 1:04Autonomous software ecosystems
- 3:52Sharing executable modules
- 5:59Real-time trust decisions
- 7:21Digital twins for prediction
- 9:30Architecture and trust scores
- 11:16Kubernetes applications
- 12:40Future research directions
- 14:06Closing remarks
related talks
40mIn AI We Trust? Securing the Future, One Agent at a Time
23mZero-Trust Architecture for Agentic AI: Securing Multi-User Access and Third-Party Integrations
27mFrom Perimeter to Silicon: Why Agentic AI Demands a New Security Model
26mTrust Is the Next Bottleneck: Why the Agentic Economy Needs Confidential
30mZero-Trust Traffic Governance for Kubernetes AI Agent Sandboxes
39mTrust, but Sandbox: Securing AI Agents in Kubernetes
24mAI Governance for developers: A practical framework for deploying autonomous coding agents safely…
28mAgent Runtime Security: Enforce, Attest, Decide