American cybersecurity technology company CrowdStrike has unveiled Falcon AI Detection and Response, a new security product designed to protect enterprise artificial intelligence systems from attacks targeting prompts and AI agent interactions.
Known as Falcon AIDR, the product extends the CrowdStrike Falcon platform into an area of AI usage that many security teams are not yet actively monitoring. It focuses on defending against prompt injection, jailbreaks and other techniques used to manipulate AI systems or extract sensitive data. Falcon AIDR is now generally available.
CrowdStrike positions AIDR as a unified system that secures AI data, models, agents, identities, infrastructure and user interactions across an organisation. The company said the product addresses growing risks linked to the rapid adoption of generative AI tools in everyday business operations.
Michael Sentonas, President of CrowdStrike, said attacks targeting prompts have become a major concern for organisations using generative AI.
“Prompt injection is a frontier security problem. Adversaries are injecting hidden instructions into GenAI tools to weaponize the very systems transforming how work gets done,” said Michael Sentonas, president of CrowdStrike. “Falcon AIDR secures every prompt, response, and agent action in real time, extending the power of the Falcon platform to the interaction layer and delivering complete protection across our customers’ AI infrastructure.”
The product is designed to secure the interaction layer where generative AI systems receive instructions and perform actions. Attackers increasingly target this layer using crafted prompts or hidden instructions to influence outputs, control AI agents or gain access to restricted data. CrowdStrike describes this layer as a new attack surface and likens prompts to a new form of malware.
As enterprises deploy chat-based AI tools and autonomous agents that connect to documents, internal systems and external APIs, new attack paths emerge. Falcon AIDR monitors prompts, responses and agent behaviour in real time, applying policy controls to block interactions that match known attack techniques or breach organisational rules.
The platform also logs how employees and AI agents interact with systems, supporting compliance requirements and post-incident investigations. CrowdStrike said AIDR draws on research into adversarial prompts and more than 180 known prompt injection techniques to detect and stop malicious activity before it can affect AI models or downstream systems.
Falcon AIDR includes real-time controls to block unsafe interactions, contain unexpected agent behaviour and prevent sensitive data, such as credentials or regulated information, from being exposed to AI tools or external providers.
The product integrates into the wider Falcon platform, which already covers endpoint, cloud workload, identity and data security. CrowdStrike said this approach gives customers a single, unified platform to secure both traditional IT environments and emerging AI systems, as organisations increasingly treat AI as a critical asset requiring direct security oversight.
