Cloudflare has launched a multi-agent AI system for its Managed Defense service, designed to automate security alert triage and investigation. The system uses deterministic code for data collection followed by specialized AI agents to analyze traffic and threat intelligence, aiming to reduce the manual workload for security analysts.
The shift from general-purpose AI to a multi-agent harness addresses critical reliability issues like hallucination and scope drift in automated security. By isolating data collection in application code and using specialized agents for analysis, Cloudflare provides a blueprint for high-stakes infrastructure providers to integrate LLMs without sacrificing deterministic accuracy. For the streaming industry, where high-volume traffic makes manual alert review impossible, this architecture offers a path toward autonomous mitigation of DDoS and application-layer threats. Watch for the rollout of the Custom Managed level in the coming quarters to see how much control Cloudflare grants organizations over these autonomous workflows.
Cloudflare has launched a multi-agent security operations system to automate the investigation of alerts within its Managed Defense service. By using specialized AI agents and deterministic code to analyze network traffic, the platform reduces manual triage and addresses reliability issues like hallucination, offering a scalable solution for high-volume infrastructure security.
The system uses specialized AI agents that run in parallel to analyze traffic, customer context, and threat intelligence, while deterministic code handles initial data collection to filter out noise.
The platform integrates specialized models including GPT-5.6 Cyber, the Clef open-source decision model on Workers AI, OpenAI Daybreak Defense Network, and Anthropic models.
A multi-agent harness addresses critical reliability issues like hallucination and scope drift by isolating data collection in application code and using specialized agents for specific analysis tasks.
Because streaming platforms face high-volume traffic that makes manual alert review impossible, this architecture provides a path toward autonomous mitigation of DDoS and application-layer threats.
Add StreamingMeme as a preferred source on Google to see more of our streaming news at the top of your Search results.
Add as preferred source
