CrowdStrike Fiddler Dreadnode

CrowdStrike is reportedly paying $495 million for two companies whose job is to watch and attack AI agents.

The deals: Fiddler AI for about $400 million and Dreadnode for about $95 million. Fiddler monitors agents in production, with pre-deployment evaluation, reasoning tracing, drift detection and hallucination monitoring. Dreadnode runs offensive testing, with 100+ jailbreak and prompt-injection techniques.

Most of the price is cash. CrowdStrike is buying capability, not a story.

My read: agent governance is moving from a feature to a budget line. When the largest endpoint security vendor pays this much for observability and red-teaming, it is telling you where it expects enterprise risk to sit in 2027.

For leaders deploying agents, the practical question is no longer whether the model is good enough. It is who can show what the agent did, why it did it, and whether anyone tried to break it first.

Companies that cannot answer those three questions will struggle to get agents past security review.

Which of those three is your organization least ready to answer today: what the agent did, why it did it, or whether it was tested adversarially?