智能边缘人工智能：威胁架构、攻击面和现实世界风险.pdf

1、#SECTORCA BlackHatEventsAgentic Edge AI:Threat Architecture,Attack Surfaces&Real-World RiskNumaan HuqSenior Threat Researcher,Trend Micro#SECTORCA BlackHatEventsWho am I?#SECTORCA BlackHatEventsAgentic AI is a software architecture that aims to solve complex tasks through autonomous agents.Each agen

2、t is typically designed to perform a specific set of functions within a particular domain and can leverage tools,such as a web client,to interact with the outside world.These tools enable agents to gather information,act upon their environment,and communicate with other systems.While these agents ar

3、e not necessarily AI-driven,they normally leverage AI,in which case they are referred to as AI agents.Agents are managed by an orchestrator the reasoning engine responsible for identifying goals,formulating a plan,and coordinating the agents workflow to achieve such goals.Agents,in turn,serve as the

4、 fundamental units that perform actions within the agentic systemWhat is Agentic AI?#SECTORCA BlackHatEventsAgentic Edge AI is an edge first architecture where a local agentic orchestrator runs compact models to perceive,reason,plan,and act in real time on-device,and can stay operational even if off

5、line,while also using the cloud to augment analytics,learning,and fleet coordination.What is Agentic Edge AI?#SECTORCA BlackHatEventsImportancePerception-Reasoning-Actuation Loop Core FunctionsPerforms sensing,cognition,learning,and action capabilities locally on edge devices.Goal-Directed AutonomyS

6、ystems can independently pursue goals without external control,enhancing adaptability and efficiency.Real-Time Decision-MakingDecisions are made instantly based on current data,enabling rapid responses to changing conditions.On-Device ProcessingData is processed locally on the device,improving speed