2026-10-31
Cloudflare Workers AI Returned Other Customers' Data in 155 Responses (GLM 5.3): What Teams Using Hosted AI Should Check
On October 26, 2026, Cloudflare published an account of a cross-request data exposure on Workers AI. From August 27 to September 9, 2026, prompts longer than 681,280 tokens sent to the GLM 5.3 model could receive responses that contained data from other customers' requests to the same model made shortly before. Cloudflare counts 155 affected responses, says no other Workers AI model was affected, and says that after excluding its own testing and the reporting researcher's, the affected requests looked like ordinary long-context use, with no evidence of deliberate exploitation. The report came through its bug bounty on September 8; the model was made private on September 9 and restored after a fix that evening. Cloudflare gives two causes: a cache pool smaller than the model's context length, and a bug in the open-source inference engine it uses. This site's take: if your team sends confidential text to any hosted AI API, ask the vendor for its incident notices, avoid sending data you could not afford to see exposed, and keep your own record of what was sent.