CrowdStrike Launches Falcon AIDR to Secure AI Prompts and Agents

CrowdStrike has announced the general availability of Falcon AI Detection and Response (AIDR), extending its Falcon platform to protect what it calls the fastest-growing attack surface in the AI era: the AI prompt and agent interaction layer.

Falcon AIDR is positioned as the industry’s first unified platform designed to secure every layer of enterprise AI—from data, models, and identities to infrastructure, agents, and human-AI interactions—across both development environments and workforce usage.

“Prompt injection is a frontier security problem,” said Michael Sentonas, President of CrowdStrike. “Attackers are embedding hidden instructions into GenAI tools to manipulate outcomes and weaponize AI systems. Falcon AIDR secures every prompt, response, and agent action in real time, extending protection to the interaction layer.”

As AI systems increasingly reason, decide, and act autonomously, adversaries are targeting this interaction layer to hijack agents, bypass safeguards, and access sensitive data. CrowdStrike describes prompts as the “new malware” in the AI age, requiring continuous, real-time protection.

Falcon AIDR delivers unified visibility, governance, and enforcement across enterprise AI usage. Key capabilities include deep visibility into employee AI usage and agent behavior, real-time blocking of prompt injection and jailbreak attacks, enforcement of AI usage policies, and protection of sensitive and regulated data before it reaches AI models or external systems.

The solution also enables developers to build secure AI applications faster by embedding safeguards directly into workflows, reducing risk without slowing innovation.

Integrated into the Falcon platform, Falcon AIDR provides organizations with a single, end-to-end security model for AI—covering both where AI runs and how it is used—helping enterprises adopt AI at scale with greater confidence and control.