Cybersecurity Research

CrowdStrike Moves to Buy Fiddler to Help Businesses Keep Watch on AI Agents

CrowdStrike · 3 Nov 2026
Key Takeaway Before you let an AI agent act on your systems, make sure you can see and review what it decides and does, not just whether its login is legitimate.

CrowdStrike has announced its intent to acquire Fiddler, a company focused on AI observability, at Fal.Con EU 2026. Fiddler's technology will join the CrowdStrike Falcon platform as Falcon Observe, which is expected to be generally available in Q4 FY27. The aim is to help businesses check AI agents before deployment and keep monitoring their behaviour afterwards.

The announcement rests on a simple point: an AI agent does not need to be hacked to cause harm. An agent working within its guardrails, using valid credentials and approved tools, can still make a poor decision or take a harmful action. CrowdStrike gives the example of an agent processing a customer refund that breaks company policy. Traditional security controls may see only legitimate activity, while AI explainability can show why the agent made that decision and where it went wrong.

CrowdStrike says safe use of AI agents needs three things: evaluating agents before they go live, understanding their decisions and actions, and monitoring them continuously. Because agents can act on their own, invoke tools and make changes across business systems, a record of what each one decided, did and changed is needed. CrowdStrike notes that auditors, regulators, board members and risk managers are increasingly asking for this kind of visibility.

AI agents CrowdStrike AI observability AI security acquisition
Running AI agents? Eight controls for securing them, mapped to the ISM ->

Summarised by CISO AI from CrowdStrike, written with Claude Sonnet 5.5. We link back to every original so you can read it yourself.