Cloudflare’s WAF emergency release introduces six new detections that target malformed Base64 payloads used in the Java4Shell vulnerability (CVE‑2026‑114938). All six rules are deployed in log‑only mode to avoid immediate false‑positive blocking, but they can be switched to a stricter action if operational data justifies it.
What Changed
The Managed Ruleset now contains six additional rule entries, each focused on a different location where a malformed Base64 string might appear:
- Header value
- Query string value
- Request body value
- Path segment
- Base64url JWT component
- WebAuthn value
Each rule is identified by a unique rule‑id (e.g., 5de9b20a45ea45d3841919ab4d718bf4) and is marked as a new detection. No previous action existed for these signatures, and the default action is now log.
Operational Impact
Running the rules in log mode means that traffic matching any of the six signatures will be recorded but not blocked. This approach reduces the risk of legitimate requests being dropped while the detection logic is still being validated in the field. Teams should therefore:
- Integrate the new log entries into existing SIEM or log‑analysis pipelines.
- Correlate the logs with other indicators of compromise to assess real‑world exploitation attempts.
- Track false‑positive rates and adjust rule actions if the data shows low noise and high confidence.
Implementation Considerations
Because the rules are part of the Cloudflare Managed Ruleset, they inherit any site‑wide rule ordering and priority settings already in place. If a practitioner decides to move a rule from log to block, the change must be made through the standard WAF configuration UI or API, specifying the desired action for the relevant rule‑id. No legacy rule identifiers are provided, so the new rules are independent of any previous signatures.
Since the release is labeled an “emergency” update, it is reasonable to assume that the detection patterns are time‑sensitive. Teams should therefore treat the new logs as a short‑term monitoring focus and revisit the rule actions after a defined evaluation window (for example, one week of log data).
Related CloudNinjas coverage: security.
What This Means For Practitioners
Practitioners should add the six rule identifiers to their alerting thresholds, review the logged events for signs of exploitation, and decide whether the risk profile warrants moving any rule to a blocking stance. Continuous evaluation of false‑positive rates will guide the appropriate balance between visibility and enforcement. Keeping an eye on the associated blog post for additional context can help fine‑tune the response without introducing unnecessary disruption.

