# Cloudflare opens early beta of AI-assisted security investigations

_Published Wednesday, October 7, 2026 at 10:35 PM EDT · Security, AI · Latest · Tier 2 — Notable_

![Cloudflare opens early beta of AI-assisted security investigations — Primary](https://blog.cloudflare.com/_emdash/api/media/file/01M32R921QMDZ1S2MHB3M7Q85H.01M32R93AJ8Q0D22GYW0163M3F.png)

Cloudflare says an early beta of its multi-agent investigation system is available in Managed Defense for eligible application-security alerts and cases. The system assembles evidence, links related alerts and recommends next steps for security analysts.

Application code collects data and fixes customer scope before models analyze it. Four specialist agents review traffic, customer history, aggregate global telemetry and threat intelligence; code checks citations against the investigation's evidence. Cloudflare uses Clef for initial scoring and approved OpenAI and Anthropic models for deeper analysis.

Analysts retain responsibility for decisions and mitigation. When evidence is insufficient, the system makes no classification or disposition recommendation.

## Sources

- [The Cloudflare Blog](https://blog.cloudflare.com/agentic-security-operations/)

---
Canonical: https://techandbusiness.org/newswire/8JHm4nlxpKimaXIoU_ZtLM
Published: 2026-10-08T02:35:06.814Z
Story chronology: 2026-10-07T16:30:36.761Z
Retrieved: 2026-10-08T04:42:31.223Z
Publisher: Tech & Business (techandbusiness.org)