Cross-request KV cache exposure in the GLM-5.3 Workers AI model (155 responses)

Cloudflare was alerted on September 8, 2026, by Gadi Evron to a tenant isolation issue in the GLM-5.3 Workers AI model, active since August 27, 2026. The issue, which caused cross-request KV cache exposure, was remediated, and the model was restored on September 9. This bug, reported to SGLang as issue #30233 on July 6, 2026, led to patches in SGLang v0.5.20, v0.5.21, and v0.5.22, with Cloudflare recommending updates to v0.5.22 or later.

This report details the specific vulnerability in GLM-5.3, unlike general security advisories, and traces its fix through SGLang versions up to v0.5.22.