An authenticated user with agent permissions can specify any AI analytics run identifier when calling the ticket summarize endpoint. The response returns the error message stored by the AI provider for that run, even if the run belongs to a ticket the agent cannot access. The flaw arises from insufficient authorization checks (CWE‑639) and privilege escalation (CWE‑862). The only data exposed is the error string; the actual ticket content remains protected.

The vulnerability is present in the Zammad helpdesk system. All versions prior to 7.1.2 are affected. Version 7.1.2 and later contain the fix introduced by the vendor.

The CVSS score of 2.3 classifies the issue as low severity. Because it requires authenticated agent access, the attack vector is internal and unlikely to be widely exploited. EPSS is not available, and the flaw is not listed in the CISA KEV catalog. The disclosed information is limited to an error string, making the overall impact modest but still worth addressing.