Cloudflare has connected its AI agent Cloudy to its security analytics tooling and added a natural-language chat interface for investigating traffic anomalies. The pitch is a familiar one in security operations: more telemetry produces the visibility needed to catch threats, but it also produces more noise than lean teams can sift through. Root cause analysis traditionally means moving between dashboards, filtering logs, and cross-referencing threat feeds while an incident is live. Cloudy is positioned as a way to compress that loop.
Adoption so far
Cloudy launched in March. According to Cloudflare, more than 54,000 users have tried it for custom rule creation, and 31% of those users deployed a rule the agent suggested. Separately, the log explainers in Cloudflare Gateway have been invoked over 30,000 times in the past month, with 80% of feedback confirming the generated summaries were insightful.
A conversational path through traffic data
Analytics dashboards assume the operator already knows what to look for and which queries will surface it. The new chat interface is built for faster root cause analysis (RCA) of traffic anomalies — the stated goal is to get from "something's wrong" to "here's the fix" in minutes, starting with a broad question and narrowing from there the way one would with a human analyst.
A session can begin by asking Cloudy to look into a recommendation from Security Analytics.

Follow-up questions can drill into specifics:
- "Focus on login endpoints only."
- "What are the top 5 IP addresses involved?"
- "Are any of these IPs known to be malicious?"
The same approach is being applied to another security workload — automated summaries of email detections — where the underlying task is translating complex data into clear, actionable output.
Extending analysis beyond your own logs
Analyzing a customer's own logs shows only part of the picture. Cloudforce One's Threat Events platform converts high-volume attack data observed across Cloudflare's network into real-time, attacker-attributed events scoped to an organization. It covers APT infrastructure, cybercrime groups, compromised devices, and volumetric DDoS activity, with interactive timelines and mappings to attacker TTPs, regions, and targeted verticals.
Cloudy is now integrated with that platform, so it can return contextual data on threats Cloudflare observes and mitigates globally — APT activity, residential proxies, ACH fraud, DDoS attacks, WAF exploits, cybercrime, and compromised devices. Queries are issued in natural language and answered concisely. Examples of what an operator can ask:
- Who is targeting my industry vertical or country?
- What are the most relevant indicators (IPs, JA3/4 hashes, ASNs, domains, URLs, SHA fingerprints) to block right now?
- How has a specific adversary progressed across the cyber kill chain over time?
- What novel new threats are threat actors using that might be used against your network next, and what insights do Cloudflare analysts know about them?
The feature lives under Cloudflare Dashboard > Security Center > Threat Intelligence. An investigation can start at a single indicator — an IP address or domain — move to the specific threat event Cloudflare observed, and then pivot to related attacks, related threats, or other activity from the same actor.

That lets an analyst trace an adversary's actions against the cyber kill chain and the MITRE ATT&CK framework, then block based on the resulting intelligence. Cloudflare describes the Threat Events platform as an evidence board and Cloudy as the sidekick that runs down each lead.
Implementation and data handling
Cloudy's threat intelligence capability is built on Cloudflare's Agents SDK running on Workers AI, which the company credits for rapid iteration and deployment during development. Cloudflare says Cloudy was not trained on any customer data; it relies on models made publicly available through Workers AI.
Availability and planned work
The new chat interface is available on all plans. Threat intelligence capabilities through Cloudy are live for Cloudforce One customers. On the roadmap: intelligent debugging for WAF rules, deeper integrations with Alerts for more contextual notifications, continued enrichment of the threat events datasets, and ways for Cloudy to visualize attacker timelines, campaign overviews, and attack graphs.



