Securing the AI PC Beyond the OS: Data, Agents, and Resilience (Oct. 13th)
Tuesday, October 13th, 2026: 1:00 PM to 2:00 PM
The discussion will be positioned around how IT and security leaders can establish a unified silicon-to-cloud defense model, combining hardware-rooted trust (Intel Core Ultra, vPro, TIEE) with AI-native endpoint protection (CrowdStrike Falcon Data Security, Guardian, Falcon for IT) to maintain visibility and control over distributed AI PC environments without sacrificing user privacy or system performance.
Virtual
Discussion Topics
- How AI PC Adoption Shifts Endpoint Risk: Moving GenAI execution and autonomous agents onto local devices introduces new data exfiltration vectors, unmonitored tool calls, and prompt-injection risks.
- Why Software-Only DLP & OS Controls Fall Short: Legacy rule-based DLP struggles with unstructured GenAI interactions, while OS-level software fails against malware-free and fileless memory attacks that dominate modern breach activity. (~79% of all breaches according to 2025 CrowdStrike Global Threat Report)
- Hardware-Accelerated Detection & Memory Scanning: Offloading memory scanning to the GPU via Intel TDT delivers 7X performance gains, while CPU telemetry catches exploits below the OS layer.
- On-Device Data Protection via the NPU: Running Falcon Data Security models locally on the Intel Core Ultra NPU enables real-time semantic classification at the point of egress—preserving privacy, eliminating cloud latency, and reducing costs.
- Hardware Guardrails for Autonomous AI Agents: Intel TIEE uses TPM 2.0 and VT-x sandboxing to block unauthorized agent actions and feed tamper-proof telemetry into Falcon Guardian.
- Out-of-Band Operational Resilience: Intel vPro AMT integration in Falcon for IT enables remote access, management, and recovery of unbootable or compromised endpoints.
Hosted by BWG