CrowdStrike’s announced agreement to acquire Fiddler AI signals a significant shift in enterprise cybersecurity, as organisations increasingly recognise that securing autonomous AI agents requires more than traditional monitoring and threat detection.
Fiddler AI, backed by Insight Partners since 2021, began its journey in 2018 with machine learning observability. The emergence of generative AI subsequently pushed the company towards agentic AI observability and intelligent governance.
At Insight Partners’ ScaleUp conference, Fiddler executive Fahad Rizqi explained the company’s evolution towards building an independent control plane for AI systems, designed to strengthen trust, accountability and operational transparency.
The fundamental challenge is that observability alone cannot guarantee security. Detecting abnormal behaviour is useful, but enterprises also need to understand its root causes, assess consequences and determine appropriate corrective actions.
Fiddler addresses these challenges through specialised evaluation models that analyse AI interactions and outputs. Its technology monitors hallucinations, potential exposure of personally identifiable information, and other performance and security indicators across AI workloads.
For CrowdStrike, the proposed acquisition represents an opportunity to extend cybersecurity capabilities into the rapidly expanding agentic AI ecosystem, where autonomous applications increasingly access enterprise data, execute workflows and influence business decisions.
However, AI security demands more than identifying suspicious outputs. Enterprises require continuous behavioural monitoring, independent model evaluation, policy enforcement, human oversight and automated containment to prevent compromised agents from triggering cascading operational failures.
The bigger industry message is clear: AI observability is evolving into AI assurance. As autonomous agents become embedded across banking, healthcare, government and enterprises, independent verification and enforceable controls will become essential pillars of digital trust and responsible AI adoption.
