CrowdStrike Falcon® Guardian
Secure AI agents at runtime
Discover, govern, and secure AI agents with runtime visibility and control. Protect AI wherever it executes, from the endpoint to cloud and SaaS environments.
New from Fal.Con 2026
AI agents don't just answer. They execute.
Computer-use agents can access identities, data, files, browsers, shells, and system settings. Acting with user permissions at machine speed, they can create an enormous agentic blast radius.
Agents live in the shadows
Security teams can’t see which agents are running, who is using them, or what risks they introduce.
Agent actions are invisible
Security teams can’t connect prompts to the actions agents execute to identify agentic misalignment.
Agents face novel runtime attacks
Prompt injection and poisoned tools can hijack agents and cause breaches at machine speed.
Secure the entire AI estate with Falcon Guardian
Falcon Guardian brings AI Detection and Response (AIDR) to the point of execution, connecting AI activity with endpoint telemetry to understand and secure autonomous AI behavior at runtime.
×
![]()
99%
Detection efficacy of prompt attacks5
100milliseconds
Latency or less for detections5
Adopt and innovate faster
Pre-built, seamless guardrails accelerate secure AI use and development compared to DIY solutions
See and secure every AI action
Unify AI visibility, governance, runtime protection, and response.
New
Discover shadow AI agents
Gain visibility into workforce AI use and endpoint agents. Uncover shadow AI to understand what’s in use, who is using it, view token consumption, and see where unmanaged risk exists.
×
![]()
×
![]()
New
See what AI agents actually do
Connect agent activity with endpoint telemetry to trace how user prompts lead to downstream processes and system actions.
New
Govern AI usage and agent access
Turn AI governance policy into enforceable controls. Govern how users interact with AI tools and define which AI agents are permitted to run.
×
![]()
×
![]()
Protect prompts and sensitive data
Apply Falcon Guardian security policies as users, applications, and agents interact with AI services. Detect and stop malicious interactions and protect sensitive data in transit without disrupting legitimate AI workflows.
New
Detect and stop threats at runtime
Detect compromised AI agents, reconstruct what happened, and determine the full blast radius across affected agents and systems. Deploy containment actions to stop threats before they spread.
×
![]()
New
Coming soon: AI gateway for Falcon Guardian
Centralize visibility and control over enterprise AI traffic. Falcon Guardian’s AI gateway capability will manage how applications and agents access AI models and services, using Falcon context across users, agents, endpoints, identities, assets, and security posture to inform policy enforcement.
See Falcon Guardian in action
Latest innovations from
Falcon Guardian launches end-to-end security for endpoint agents and announces new AI gateway capability
New Game: Agents of Chaos. Infiltrate a shadow organization, stop the threat, and play for a chance to win up to $100,000.*
Featured Resources
Falcon Guardian FAQs
CrowdStrike Falcon® Guardian is our flagship AI detection and response (AIDR) solution. It helps organizations discover and govern AI use, secure autonomous AI agents at runtime, protect sensitive data, and detect and respond to AI threats.
AI Detection and Response (AIDR) is a cybersecurity solution category for discovering, governing, and securing AI systems and activity. AIDR solutions extend protection from AI interactions to runtime execution, helping organizations detect and stop AI threats as they occur.
AI agents can use tools, access data, and take actions with user permissions at machine speed. Runtime security provides visibility and control as those actions execute, helping detect and stop threats before they lead to a breach.
Falcon Guardian correlates AI agent activity with endpoint telemetry to connect user prompts to downstream system actions. This runtime context helps security teams understand agent behavior, investigate threats and their blast radius, and stop attacks before they become breaches.
Shadow AI is AI use or deployment that operates outside an organization’s visibility or control Falcon Guardian discovers AI use and agents on endpoints, and across cloud and SaaS environments,helping security teams understand activity, enforce policy, and reduce unmanaged AI risk.
Yes. Falcon Guardian protects workforce use of AI including runtime security for AI agents used on the endpoint. It also helps organizations secure homegrown AI agents, factories, and workloads from runtime threats like prompt injection attacks.
Falcon Guardian identifies sensitive data in AI interactions and applies runtime controls to redact, encrypt, or block it before exposure. This helps organizations enforce data protection policies while keeping legitimate AI workflows moving.
Falcon Guardian inspects AI interactions at runtime to detect direct and indirect prompt injection attacks spanning 200+ techniques. It can block malicious interactions before they manipulate AI agents, expose sensitive data, or trigger unsafe actions.
Traditional endpoint security detects and stops threats based on host-level activity. Falcon Guardian adds AI and agent-specific context, connecting prompts and agent activity to downstream endpoint actions so teams can understand, investigate, and stop threats as AI executes. Traditional endpoint security only sees one aspect of this AI behavior, the host-level activity.
1CrowdStrike Taxonomy of Prompt Injection
2IBM Security & Ponemon Institute: Cost of a Data Breach Report 2025.
3The CrowdStrike Falcon® Adversary OverWatch™ threat hunting team
4KPMG Global AI Pulse, Q1 2026
5Performance metrics are based on results from internal benchmark testing.
*NO PURCHASE NECESSARY TO PARTICIPATE OR WIN. Challenge begins 8/31/26 at 5:00 p.m. PT and ends 9/29/26 at 11:59 p.m. PT. Must be at least age of majority in jurisdiction of residence to participate. Void in select jurisdictions & where prohibited. See full Official Rule, which govern the challenge, for complete details.
